{ "schema_version": 1, "date_from": "2026-01-01", "date_to": "2026-08-24", "unique_new_candidates": 3906, "papers": [ { "arxiv_id": "2608.21360", "title": "OmniAssistBench: Assistant-style Interaction Benchmark for Omni-LLMs", "published": "2026-08-21T17:59:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21356", "title": "AI with Authority, from Application to Silicon", "published": "2026-08-21T17:59:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21341", "title": "Natural-Language Workflows Are Not Software Yet: Artifact-Driven Compilation for Reliable Agent Execution", "published": "2026-08-21T17:47:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21311", "title": "AI-to-AI Code Reviews of GitHub Pull Requests", "published": "2026-08-21T17:17:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21310", "title": "Beyond Fault Localization: A Trajectory-Level Study of LLM Agents for Microservice Root Cause Analysis", "published": "2026-08-21T17:13:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21308", "title": "Rethinking Expressivity and Efficiency in Test-Time Training", "published": "2026-08-21T17:12:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21305", "title": "Re$^3$Cap: Retrieval-Guided Refinement for Image Captioning Enhancement via Reinforcement Learning", "published": "2026-08-21T17:07:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21274", "title": "Recommendation Quality and the Concentration of Consumption: Experimental Evidence from Netflix", "published": "2026-08-21T16:30:18Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-netflix", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21265", "title": "Memory Augmentation Unlocks Efficient Chain-of-Thought Reasoning", "published": "2026-08-21T16:22:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21249", "title": "Benchmarking Patent Drafting from Inventor-Style Disclosures", "published": "2026-08-21T16:00:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21247", "title": "Just Noticeable Difference Modeling for Token Compression in Vision-Language-Action Models", "published": "2026-08-21T15:59:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21244", "title": "A VLM Answer Is Not an Anomaly Score: Rank Compression in Training-Free Video Anomaly Detection", "published": "2026-08-21T15:56:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21223", "title": "Event-triggered Implicit Perturbation for Zeroth-Order Fine-Tuning of Spiking Transformers", "published": "2026-08-21T15:32:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21209", "title": "Personalized Privacy Control in LLMs via Attention Head Intervention", "published": "2026-08-21T15:22:20Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21208", "title": "Specification Portability Across LLM Development Agents: Cross-Agent Compatibility in Specification-Driven Software Migration", "published": "2026-08-21T15:21:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21199", "title": "Tydra: An Efficient Hybrid Model for Tabular Data", "published": "2026-08-21T15:15:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21186", "title": "A Neurosymbolic Approach for Constructing Planning Domain Models from Clinical Narratives", "published": "2026-08-21T14:57:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21180", "title": "Toward Vision Language Model-based Assessment of Clinical Quality and Usability of LGE-MR Images for Cardiac Ablation Planning", "published": "2026-08-21T14:51:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21172", "title": "Thermo-FL: Thermal-Aware Robust Federated Fine-Tuning of Large Language Models for Edge AI", "published": "2026-08-21T14:41:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21170", "title": "Is Visual Prompting All You Need? Studying VLM Spatial Reasoning under Progressive Visual Scaffolds", "published": "2026-08-21T14:40:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21156", "title": "Graph Engineering in the Era of LLM Agents: From Individual Intelligence to System Intelligence", "published": "2026-08-21T14:27:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21142", "title": "COEC: Calibrated Orthogonal-Equivalence Compensation for Structured Pruning of Large Language Models", "published": "2026-08-21T14:20:49Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21140", "title": "A Modular Agent for Reliable and Auditable Spatial Relation Verification in CT Scans", "published": "2026-08-21T14:16:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21136", "title": "Stream3Dv2: Geometric-Semantic Fusion Enhanced Streaming Zero-Shot 3D Scene Understanding", "published": "2026-08-21T14:13:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21134", "title": "Llama-Mobile: Efficient 2.7-Bit Quantization of VLMs", "published": "2026-08-21T14:10:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21107", "title": "Large Language Models at the Intersection of Software Engineering and Software Security:An Evidence-Centered Structured Survey and Research Agenda", "published": "2026-08-21T13:54:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21101", "title": "ClawSentry: A Progressive Multi-Tier Security Monitor for Safeguarding Autonomous LLM Agents", "published": "2026-08-21T13:47:51Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-agent", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.21095", "title": "Trustworthy RAG: An Evaluation Agent for Detecting Misinformation and Knowledge Poisoning in Generative AI Systems", "published": "2026-08-21T13:42:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21088", "title": "When the Feature Pool Goes Algorithmic: Extending Mufwene's Ecology of Language Evolution to LLM-Mediated Exposure", "published": "2026-08-21T13:32:26Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "post-training", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21079", "title": "Causal Modeling of Adverse Pregnancy Outcomes via Adaptive LLM Proposals", "published": "2026-08-21T13:24:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21074", "title": "PromptResponse: Optimizing Prompts for LLM Coding Tasks", "published": "2026-08-21T13:16:48Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21069", "title": "Spike-Killer: Evidence-Gated LLM Assistance for Safe Performance Diagnosis on a Real Windows Workstation", "published": "2026-08-21T13:11:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21049", "title": "$Z^2$-ACT: End-to-End Verifiable Agentic Intent Control for Open 6G RAN", "published": "2026-08-21T12:44:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21030", "title": "COMET: Contrastive Motion-Enhanced Temporal Reasoning for Video Multimodal Large Language Models", "published": "2026-08-21T12:28:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "multimodal-llm", "vision-language" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.21027", "title": "Don't Solve, Just Compare: Tiny Advisors for Runtime Intervention in LLM Agents", "published": "2026-08-21T12:20:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21022", "title": "Recognition-Conditioned Reasoning: A Training-Free Multimodal-LLM Pipeline for Fine-Grained Micro-Action Understanding", "published": "2026-08-21T12:10:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21021", "title": "Free-Text Evaluation of LLMs for 5G Domain Knowledge and Fault Analysis using LLM-as-Judge", "published": "2026-08-21T12:09:51Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-architecture", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21019", "title": "Target-Aware Calibration Data Selection for Preserving Uncertainty in Quantized Language Models", "published": "2026-08-21T12:07:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.21012", "title": "From a Static Multi-Level Small Semantic Codebook to a Dynamic Single-Level Large Semantic Codebook for Generative Recommendation", "published": "2026-08-21T11:58:44Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "generative-recommendation", "production-evidence", "recsys-general" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B01", "plan_reason": "8 月工业生成推荐与多模态 — completed" }, { "arxiv_id": "2608.20999", "title": "Latent Ordinal Evidence, Misaligned Outputs: Inference-Time Ordinal Lens Alignment for Multimodal LLMs", "published": "2026-08-21T11:33:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20988", "title": "Jacobian-guided Noise Injection for Quantization Robustness in Large Language Models", "published": "2026-08-21T11:16:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20984", "title": "MigrationNarrate: A Dataset for Detection of Migration Narratives in YouTube Videos", "published": "2026-08-21T11:10:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20975", "title": "Belief Without Behavior: Measuring the Translation of Theory of Mind into Coordinated Social Action in Vision-Language Models", "published": "2026-08-21T10:55:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20963", "title": "Vibe Coding and Web Application Security: A Twin-Prompt Study", "published": "2026-08-21T10:34:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20961", "title": "TreeWY: Speculative Verification for Gated DeltaNet Hybrids", "published": "2026-08-21T10:31:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20953", "title": "Quantization-Aware Healing: A Practical Recipe for Recovering Compressed, 4-Bit LLMs", "published": "2026-08-21T10:19:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20932", "title": "OccluRank: Controllable Occlusion-Aware Layout-to-Image Generation by Adding Just an Ordinal Rank", "published": "2026-08-21T09:53:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20927", "title": "MentorPulse: Refreshing Cross-Model Latent Guidance for Long-Form Generation", "published": "2026-08-21T09:49:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20920", "title": "ForeDreamer: A Self-Evolving Dual-Agent Memory Architecture for Future Event Prediction", "published": "2026-08-21T09:38:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20903", "title": "Generation of Web Apps with Agentic IDEs: An Empirical Assessment", "published": "2026-08-21T09:21:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20890", "title": "A Collaborative Multi-Modality Interaction for VLA-based End-to-End Autonomous Driving", "published": "2026-08-21T09:06:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20868", "title": "Identify, Locate, Link: End-to-End Key-Value Extraction from Document Images", "published": "2026-08-21T08:36:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20851", "title": "BC-Bench: Evaluating Agentic Engineering in a Domain-Specific Language for ERP", "published": "2026-08-21T08:16:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20844", "title": "TRACE: Agentic Catalog Enrichment with Multi-source Evidence Grounding", "published": "2026-08-21T08:08:27Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-agent", "multi-agent", "web-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.20831", "title": "STAR-OPD: Structured Aspect-Cascade-Aware On-Policy Reward Distillation for ABSA Quadruple Extraction", "published": "2026-08-21T07:50:00Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20830", "title": "Fine-tuning LLMs for Tourist Trajectory Prediction using Field Experiment Data", "published": "2026-08-21T07:49:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20814", "title": "Enhancing Localized Reasoning for Long Video Understanding via Efficient Segment-to-Video Supervision", "published": "2026-08-21T07:34:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20801", "title": "Profiling What Matters: Context-Aware Item Profiles from Large-Scale Metadata for LLM Recommenders", "published": "2026-08-21T07:20:06Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20791", "title": "CertVLA: Certified Defense against Physical Visual Attacks for Vision-Language-Action Models", "published": "2026-08-21T07:09:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20786", "title": "Structure for Reading, Prose for Writing: Asymmetric Structural Conditioning in Multi-Agent Document Authoring", "published": "2026-08-21T06:57:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20777", "title": "Tree-of-Concerns: Hierarchical Multi-Agent Debate for Unstated-Limitation Extraction in Scientific Critique", "published": "2026-08-21T06:41:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20771", "title": "CAS: Conformalized Agentic Search via Adaptive Retrieval and Policy Weighting", "published": "2026-08-21T06:29:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20763", "title": "CARD: Diagnosing Belief to Action Routing Failures in Vision Language Models", "published": "2026-08-21T05:53:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20756", "title": "Vis-Poison: Poisoning Visual Knowledge in Multimodal Retrieval-Augmented Generation", "published": "2026-08-21T05:34:48Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20749", "title": "Identity-Preserving Text-to-Video Generation via Agentic Enhancement and Semantic Repair", "published": "2026-08-21T05:19:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20729", "title": "Calibrating Criterion Revision in LLM Agents: Failure Modes and a Trace-Anchored Protocol", "published": "2026-08-21T04:21:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20720", "title": "AffordAny: Open-World 3D Affordance Grounding from Monocular RGB Images via Vision-Language-Guided Geometric Reasoning", "published": "2026-08-21T03:55:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20713", "title": "AGIDefect-4K: A Richly Annotated Dataset for AI-Generated Image Defect Detection, Localization and Explanation", "published": "2026-08-21T03:40:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20711", "title": "AsmEvo: Agentic Assembly-Level Optimization of AMD GPU Kernels with Functional Equivalence Verification", "published": "2026-08-21T03:34:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20699", "title": "ArtiMo: Agent-Driven Articulated Mesh Animation", "published": "2026-08-21T03:08:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20688", "title": "VortexChat: An agentic framework for autonomous multi-objective integrated photonic design", "published": "2026-08-21T02:50:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20686", "title": "CDRL: Certification-Driven Reinforcement Learning for Neutrino Flavor Model Discovery", "published": "2026-08-21T02:49:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20670", "title": "Why2Speak: Faithful Reasoning for Abstaining Action Policies", "published": "2026-08-21T02:00:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20668", "title": "Lightweight Adaptive ReduNet via Hyperspherical Manifold Learning", "published": "2026-08-21T01:58:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20664", "title": "DreamBench-SWE: A Multi-Session Memory-Hygiene Benchmark for Software Agents", "published": "2026-08-21T01:48:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20640", "title": "One Hierarchy, Two Systems: Semantic Product IDs for Discovery-Surface Ranking and Search-Page Query Reformulation", "published": "2026-08-21T00:27:32Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20637", "title": "ARQ: Agentic CodeQL Query Refinement for C/C++ Vulnerability Detection", "published": "2026-08-21T00:20:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20634", "title": "AgentMercury: Your Agent Can Synthesize Verifiable Environments for Business Scenarios at scale", "published": "2026-08-21T00:15:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20631", "title": "Weighted Memory Tree: Remembering What Matters for Long-Horizon LLM Agents", "published": "2026-08-21T00:14:38Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agent-planning", "llm-agent", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.20622", "title": "Applying Anthropic Primitives at Large Enterprises: Harness Paradigm for Knowledge Work", "published": "2026-08-20T23:44:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20621", "title": "RECOUNT: Reference-guided Counting with Synthetic Visual Exemplars", "published": "2026-08-20T23:41:30Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20617", "title": "Dual-Cache Latent Space Communication between Heterogeneous Language Models", "published": "2026-08-20T23:35:36Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-agent", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20614", "title": "Evaluating Skills, Not Just Agents: Agentic Continuous Evaluation of Skills", "published": "2026-08-20T23:26:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20597", "title": "Testing and Evaluation of Agentic AI Systems In Military Command and Control", "published": "2026-08-20T22:31:23Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20571", "title": "AutoMOOSE: Use Case and Logical Views of Agentic Phase-Field Simulation Software", "published": "2026-08-20T21:10:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20566", "title": "AgentDecarbonizer: Carbon-Aware Execution for AI Agents", "published": "2026-08-20T21:05:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20564", "title": "Consilience: Conformally Calibrated Communication Control for Hidden-Profile Multi-Agent Reasoning", "published": "2026-08-20T20:57:25Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20563", "title": "Beyond End-to-End Success: Diagnosing Failures in Long-Horizon Security LLM Agents", "published": "2026-08-20T20:55:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20554", "title": "aiXamine: Unified Black-Box Evaluation of Cross-Dimensional Trade-offs in LLM Safety, Security, and Privacy", "published": "2026-08-20T20:33:35Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20549", "title": "Volumetric Radiology AI in the Era of Multimodal Large Language Models", "published": "2026-08-20T20:23:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20519", "title": "An integrated diffusion-weighted imaging processing and interpretation platform for MR-guided radiotherapy", "published": "2026-08-20T19:33:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20518", "title": "FL-MAESTRO: Multi-Agent LLM Orchestration for Resource-Constrained Federated Learning", "published": "2026-08-20T19:27:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20494", "title": "Towards Traffic Modelling of Multi-Agent Systems: The Role of Coordination Topology", "published": "2026-08-20T18:30:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20492", "title": "Annotations as Rollouts: Efficient and Scalable Reinforcement Learning for Video MLLMs", "published": "2026-08-20T18:28:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20485", "title": "Terminal Agents: A Survey of AI Agents in Command-Line Environments", "published": "2026-08-20T18:16:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20481", "title": "AEGIS: Preventing Cross-Domain Resource Abuse in MCP", "published": "2026-08-20T18:12:20Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20473", "title": "Aggregating Visual Information with Optimal Transport for VideoLM Token Compression", "published": "2026-08-20T18:05:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20331", "title": "G-CARL: Grounded Checklist-Aligned Reward Learning for Patient-Oriented Medical Report Interpretation", "published": "2026-08-20T17:59:46Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "llm-rl", "multimodal-llm", "post-training", "preference-optimization", "reward-model", "vision-language" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.20320", "title": "An Agentic Approach for Active Data Collection, Travel Behavior Modeling, and Weather-Sensitive Demand Prediction", "published": "2026-08-20T17:57:42Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "multi-agent", "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20318", "title": "AI4AI-Bench: Benchmarking LLM Agents in Algorithmic Design for Recursive Self-Improvement", "published": "2026-08-20T17:56:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20315", "title": "Explainable Transformer Models for Clinical Prediction Tasks on Structured Electronic Health Records", "published": "2026-08-20T17:54:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "pretraining-data", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.20314", "title": "MidTool: Mid-training Data Synthesis for Agentic Tool Use", "published": "2026-08-20T17:53:59Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agentic-rl", "llm-agent", "software-agent", "tool-agent", "web-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.20281", "title": "Inject, Align, Recover: Staged Post-Training for Retrieval-Free Document Knowledge Internalization", "published": "2026-08-20T17:14:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20274", "title": "Break It Down, Pass It On: Cross-Task Skill Transfer in LLM Agents", "published": "2026-08-20T17:12:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20210", "title": "Daedalus-150M: A Convolution-Attention Hybrid Designed for CPU Inference", "published": "2026-08-20T16:09:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20204", "title": "ContractScrub: A benchmark for final review of legal contracts", "published": "2026-08-20T16:01:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20201", "title": "The Third Restructuring of Software Form: From the Three-Tier Architecture to Storage, Models, and Agents", "published": "2026-08-20T15:59:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20195", "title": "From Agent Behaviour to Agent-Friendly Documentation: An Empirical Study of How Coding Agents Discover, Read, and Write Technical Documentation", "published": "2026-08-20T15:51:54Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20169", "title": "Task-CoEvolve: Efficient Harness Optimization via Adaptive Validation Task Selection", "published": "2026-08-20T15:24:54Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20157", "title": "G3Ego: Gaze-Guided Graphs for Egocentric Action Understanding", "published": "2026-08-20T15:15:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20141", "title": "DPC-Net: Dual-Prior Collaborative Network for All-in-One Image Restoration", "published": "2026-08-20T15:04:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20129", "title": "Multi-Agent Orchestration with the Common-Sense Reasoning Capabilities of LLMs for Autonomous Driving", "published": "2026-08-20T14:56:15Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "llm-rl", "multi-agent", "multimodal-llm", "reward-model", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20123", "title": "Discrete Diffusion Inference-Time Control with Nested Sequential Monte Carlo", "published": "2026-08-20T14:52:17Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20107", "title": "BeyondMasks: Evaluating Causal and Physical Consistency in Video Object Removal", "published": "2026-08-20T14:37:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20106", "title": "OenoBench: A Wine-Domain Benchmark for Knowledge-Grounded Evaluation of Large Language Models", "published": "2026-08-20T14:37:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20099", "title": "Reward-Guided Autoregressive Graph Generation for Efficient Multi-Agent Communication Topology Design", "published": "2026-08-20T14:32:01Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-agent", "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20061", "title": "Let's Scale Step by Step: Compute-Efficient Hyperparameter Transfer for Large-Scale Mixture-of-Experts", "published": "2026-08-20T13:57:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20026", "title": "From Street View Imagery to Street Quality Indicators: Vision Language Inference for the Suburban 15-minute City", "published": "2026-08-20T13:34:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19998", "title": "SCoRD: Semantic-Assisted Continual Retriever-Reranker Distillation for LLM-Based Recommendation", "published": "2026-08-20T13:11:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19993", "title": "Optimal Skill Selection for LLM Agents with Provable Bicriteria Guarantees", "published": "2026-08-20T13:08:17Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19974", "title": "ReguSim: Evaluating LLM Agent Rule Grounding in Financial Compliance", "published": "2026-08-20T12:49:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19971", "title": "Robust Incomplete Multimodal Sentiment Analysis via Iterative Proxy Correction", "published": "2026-08-20T12:45:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19965", "title": "Flow Matching Meets 3D Curvilinear Structure Segmentation in Medical Imaging", "published": "2026-08-20T12:35:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19957", "title": "Natural Language Code Retrieval for 1C:Enterprise: An Open Benchmark and Efficient Bi-Encoder", "published": "2026-08-20T12:24:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20438", "title": "Peer-Voted LLM-Agent Stress Tests Find Feed-Induced Lexical Convergence but No Reliable Matched-Exposure Advantage for Distributed Sources", "published": "2026-08-20T12:01:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19920", "title": "Learning how to Forget: Fine-tuning for Long-Context Sparse Attention", "published": "2026-08-20T11:37:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19901", "title": "MaliciousSkillBench: A Comprehensive Benchmark for Malicious Agent Skill Detection", "published": "2026-08-20T11:13:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19893", "title": "Interrupting the Loop: Periodic Subject Changes Raise Judged Surprise and Connection in Base Language Models", "published": "2026-08-20T11:01:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19880", "title": "EnvHarness: Awakening Static Worlds for Agent Learning", "published": "2026-08-20T10:42:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19871", "title": "DIFFCZSL: Compositional Zero-Shot Learning Regularized by Diffusion Representations", "published": "2026-08-20T10:30:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19861", "title": "PolicyGuide: From Guarding One Action to Guiding the Whole Workflow for Policy-Compliant LLM Agents", "published": "2026-08-20T10:13:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19854", "title": "Repo0: Design-Driven Zero-to-All Code Generation", "published": "2026-08-20T10:03:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19842", "title": "SAPO: Single-Rollout Autoregressive Policy Optimization for Agentic Reinforcement Learning", "published": "2026-08-20T09:43:47Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agentic-rl", "llm-agent" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B10", "plan_reason": "Agentic RL 与长时序 credit assignment — completed" }, { "arxiv_id": "2608.19836", "title": "Adaptive Probabilistic Shielding by Learning MDPs for Safe Reinforcement Learning", "published": "2026-08-20T09:37:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19833", "title": "Do Sequential Recommendation Benchmarks Really Require Higher-Order Sequence Modelling?", "published": "2026-08-20T09:33:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19825", "title": "Towards Clinically Faithful Medical Image Captioning via Enhanced Vision-Language Alignment", "published": "2026-08-20T09:25:30Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19803", "title": "MileGPO: Milestone Inference with Local Evidence for Graph-Based Policy Optimization of Long-Horizon LLM Agents", "published": "2026-08-20T08:58:27Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-agent", "llm-rl", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.19802", "title": "Stopping and Routing LLM Judge Panels", "published": "2026-08-20T08:58:00Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19799", "title": "SWE-bench Science: Can Coding Agents Resolve Engineering Tasks in Science?", "published": "2026-08-20T08:53:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19794", "title": "Towards general embodied intelligence: integrating large language models, knowledge bases, and reasoning capabilities to build the next generation of AI agents", "published": "2026-08-20T08:45:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19784", "title": "PRAXIS: Graph-Grounded Tacit Knowledge for Domain Code Generation", "published": "2026-08-20T08:28:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19760", "title": "Credit Without Ground Truth: Auditing Step-Level Credit Assignment in LLM Agents Against Executed Replay", "published": "2026-08-20T08:04:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19758", "title": "FlashPrefill V2: Block-Sparse Prefill Attention for Long-Context LLM Serving", "published": "2026-08-20T08:02:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19748", "title": "Truncate Bad, Upweight Good: BoN-Style Distillation via Rank-Based Classification", "published": "2026-08-20T07:49:21Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19746", "title": "PersonalBench: Measuring the Authorship Gap in LLM Personalization", "published": "2026-08-20T07:48:20Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19741", "title": "One Success Isn't Reliability: Thinkingbox, a Sandbox and Benchmark for Agents in Stateful Business Workflows", "published": "2026-08-20T07:37:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19739", "title": "Question-Guided Evidence Acquisition for Multimodal Visual Question Answering", "published": "2026-08-20T07:37:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19737", "title": "TempJail: Temporal Jailbreak Attack against Large Vision-Language Models via Subtitle Scheduling", "published": "2026-08-20T07:37:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19729", "title": "SafeBranch: Branch-Pair Safety Alignment for Embodied Agents", "published": "2026-08-20T07:29:07Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19726", "title": "Projector Is All You Train", "published": "2026-08-20T07:23:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19723", "title": "StreamSoccer: Event-Driven Memory for Streaming Soccer Commentary", "published": "2026-08-20T07:19:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19703", "title": "Loreley: Repository-Scale Program Evolution with Quality-Diversity Search", "published": "2026-08-20T07:04:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20434", "title": "An LLM agent for end-to-end computational materials discovery", "published": "2026-08-20T06:23:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19670", "title": "The Asymmetric Harms of LLM Compression", "published": "2026-08-20T06:06:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19665", "title": "Training-Free LLM-Based Recommendation with Post-LLM Item Refinement Using Collaborative Signals", "published": "2026-08-20T06:01:59Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19662", "title": "ReCache: Efficient KV Cache Reuse and Compression for Tool-Augmented LLM Agents", "published": "2026-08-20T05:57:24Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-memory", "efficient-inference", "llm-agent", "model-compression", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.19652", "title": "Can Agent Memory Systems Track Evolving State?", "published": "2026-08-20T05:41:23Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19621", "title": "Mitigating Identity Essentialism in LLM Agents with Longitudinal Life Trajectories", "published": "2026-08-20T04:13:11Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-agent", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19613", "title": "What Matters for Latent Actions in Robot Learning", "published": "2026-08-20T03:54:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19598", "title": "PEA-DPO: Perception-Enhanced Alignment Direct Preference Optimization for MLLMs Alignment", "published": "2026-08-20T03:35:07Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "preference-optimization", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19589", "title": "OrthoSkillVLA: Continual Skill Learning via Gradient-Informed Skill Subspace Adaptation", "published": "2026-08-20T03:10:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20432", "title": "ProofJudge: Tool-Grounded LLM Evaluation of Formal Proof Quality in Mathlib", "published": "2026-08-20T02:39:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19564", "title": "Remember, Verify, or Ask? Cross-Family Evaluation of Memory Commitment in LLM Agents", "published": "2026-08-20T02:11:03Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "llm-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.19558", "title": "Reliable Financial Named Entity Recognition under Domain Shift", "published": "2026-08-20T01:55:41Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19553", "title": "Where Grounding Accuracy Lives on the IoU Curve: Label-Free Inference-Time Boundary Refinement", "published": "2026-08-20T01:40:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19545", "title": "Two-sided receptivity to conversational AI agents in online dating: Bilingual survey data from Fledge.Love", "published": "2026-08-20T01:25:03Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19535", "title": "From Retrieved Context to Runtime Control: Adaptive Compression for Edge-based RAG", "published": "2026-08-20T01:13:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19529", "title": "When Machines Speak: A Unified Generative Framework for Integrating Machine-Native Symbols into Pretrained Large Language Models", "published": "2026-08-20T00:54:36Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19527", "title": "Does Listening Matter? Backchanneling and Nodding in AI Clone", "published": "2026-08-20T00:53:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19509", "title": "An Agentic RAG and Evaluation Framework for Assurance Case Generation: Industrial Use Case for the EU Cyber Resilience Act Compliance", "published": "2026-08-19T23:54:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19492", "title": "Beyond Multimodal Alignment: Certifying Physical Language through Response Substitution and Ordered Execution", "published": "2026-08-19T23:04:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19490", "title": "Fine-Tuning VLAs with Self-Demonstrated Generative Control for Multi-Task Manipulation", "published": "2026-08-19T23:02:07Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "opd", "pretraining-data", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19487", "title": "Accelerated Genetic Programming Hyper-Heuristics for Simulation-Based Scheduling via Agentic AI", "published": "2026-08-19T22:48:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19408", "title": "Beyond Imitation: Filtering On-Policy Distillation by Reasoning Progress", "published": "2026-08-19T19:45:32Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "on-policy-distillation", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B08", "plan_reason": "OPD 与多教师/过程蒸馏 — completed" }, { "arxiv_id": "2608.19395", "title": "HYDRA: A Heterogeneous Chiplet DSE Framework for Serving Dynamic Hybrid LLM Workloads", "published": "2026-08-19T19:26:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "long-context", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.19389", "title": "Concentrated Liquidity Provision: a Reinforcement Learning Perspective", "published": "2026-08-19T19:08:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19385", "title": "Beyond Recognition: Compact Multi-Domain Arabic Manuscript HTR with Candidate-Selection Analysis and Evidence-Preserving Review", "published": "2026-08-19T19:04:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19380", "title": "CAViAR: A Causal Video Dataset for Fine-Grained Accident Reasoning in Real-World Scenarios", "published": "2026-08-19T18:51:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19376", "title": "Does Marginal Coverage Guarantee Class-Conditional Safety for Zero-Shot VLMs Under Shift?", "published": "2026-08-19T18:42:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20427", "title": "BF1: A Causal Dyadic Sparse-Attention Retrofit for Efficient Long-Context Transformers", "published": "2026-08-19T18:35:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19355", "title": "GRACE: Grounded Reasoning via Adapter Composition and Evidence-Aware Calibration for Educational Visual Question Answering", "published": "2026-08-19T18:20:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19197", "title": "SPADE: Self-Play in Adaptive Synthetic Executable Environments", "published": "2026-08-19T17:58:56Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-agent", "multi-agent", "reward-model", "tool-agent" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B10", "plan_reason": "Agentic RL 与长时序 credit assignment — completed" }, { "arxiv_id": "2608.19303", "title": "Outcome Monitors: Recovery Affordances for Silent Tool Failures", "published": "2026-08-19T17:35:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19075", "title": "ReWEIGH the Evidence: Calibrating Token-Level Ordinal Visual Evidence to Mitigate Hallucinations in Large Vision-Language Models", "published": "2026-08-19T16:23:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19072", "title": "What is Missing from AI Post-Training AI: An Empirical Analysis", "published": "2026-08-19T16:17:39Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19049", "title": "Multi-Agent Off-Policy Deep Reinforcement Learning for Smart Campus Coverage", "published": "2026-08-19T15:41:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19025", "title": "Self-prompting and cross-model consensus enable reproducible data extraction from scientific literature with large language models", "published": "2026-08-19T15:20:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19009", "title": "Grading the Graders: Verification Autonomy Levels (L0-L5) for LLM Reasoning", "published": "2026-08-19T15:10:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19002", "title": "A Theory of Post-hoc Debate Judgement", "published": "2026-08-19T15:03:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19000", "title": "Mise-en-Scène: Implicit Layout Emergence in Diffusion Transformers for Human-AI Design Co-Creation", "published": "2026-08-19T15:02:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18993", "title": "ForeSightGuide: An Anticipatory Framework toward Accurate and Low-Redundancy Guidance for the Visually Impaired", "published": "2026-08-19T14:58:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18984", "title": "Uncertainty-Aware Art-Historical Dating with Vision-Language Models", "published": "2026-08-19T14:50:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18952", "title": "rEDMRec: Distilling Large Language Model Reasoning into an Editable Experience Memory for Recommendation", "published": "2026-08-19T14:17:34Z", "tracks": [ "agent", "foundation-model", "post-training", "recommendation" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "agent-memory", "llm-agent", "llm-recommendation", "long-context", "model-compression", "preference-optimization", "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18940", "title": "Training Chemical Plausibility-Aware Large Language Models for Single-Step Retrosynthesis", "published": "2026-08-19T14:08:05Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18938", "title": "Breaking the weakest link to evade vision language models", "published": "2026-08-19T14:06:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18937", "title": "MedUAG: Unified Understanding and Generation for Medical Multimodal Models", "published": "2026-08-19T14:05:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18936", "title": "Graphical Design of Interpretable Architectures", "published": "2026-08-19T14:04:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18933", "title": "SkillForge: Self-Distilling Agents for Project-Specific Issue Resolution", "published": "2026-08-19T14:01:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18931", "title": "Test-Time Scaling in the Wild: Why Exploitation, Not Exploration, Is the Bottleneck", "published": "2026-08-19T13:59:53Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18884", "title": "Training-Free Inference-Time Self-Reflection and Cost-Bounded Early Stopping for Large Language Models", "published": "2026-08-19T13:09:57Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agentic-rl", "llm-agent", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.18846", "title": "ORBITER: Conflict-Aware Decision-Making for Agentic Last-Mile Delivery", "published": "2026-08-19T12:13:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18836", "title": "Verifiable abstention makes AI leak diagnosis accountable in water distribution networks", "published": "2026-08-19T11:54:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18833", "title": "EVADE: Evidence-Verified Agentic Diagnosis with Escape", "published": "2026-08-19T11:51:24Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18827", "title": "MLREF: Efficient Module Reuse for Reward Design in Reinforcement Learning via Large Language Models", "published": "2026-08-19T11:36:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18825", "title": "Understanding Multilingual Medical ASR Adaptation Through Layer-Wise Analysis", "published": "2026-08-19T11:29:30Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19285", "title": "Clustering and Token Denoising for Faster and More Robust VLMs", "published": "2026-08-19T11:20:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18779", "title": "SIDScope: A Diagnostic Resource for Semantic-ID Interfaces in Generative Recommendation", "published": "2026-08-19T10:35:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18744", "title": "Metrics That Write Themselves: Evolving an Evaluator from Its Own Blind Spots", "published": "2026-08-19T09:55:17Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18740", "title": "A Multi-Agent Platform for Automated Enterprise Analytics and Insight Generation", "published": "2026-08-19T09:49:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18736", "title": "FedLNS: Leverage LayerNorm Signature Modeling to Mitigate Adversarial Manipulation in Federated LLMs", "published": "2026-08-19T09:43:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18734", "title": "CL4D: Contrastive Language-4D Pretraining for Vision-Language Reasoning in Dynamic Scenes", "published": "2026-08-19T09:37:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18733", "title": "Flama: a Python framework for development and deployment of production-ready APIs, machine learning, and LLM services", "published": "2026-08-19T09:37:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "model-compression", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.18727", "title": "Visual-Aware Representation of Web Pages for Machine Learning Applications", "published": "2026-08-19T09:28:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18719", "title": "Competence, Not Accuracy: A Diagnostic for Reference-Free Judge Gates in Skill Optimization", "published": "2026-08-19T09:19:23Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18715", "title": "The Impact of CutMix on Reliability and Robustness in Semantic Segmentation", "published": "2026-08-19T09:16:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18704", "title": "MemFuse: Multi-Source Memory Fusion from Fragmented Observations", "published": "2026-08-19T09:04:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18696", "title": "Impact of Iterative Fine-Tuning on Transcription Accuracy in Complex Historical Sanskrit Manuscripts", "published": "2026-08-19T08:50:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18689", "title": "Aslema at NADI 2026: Augmentation through Fewshot for SLU", "published": "2026-08-19T08:41:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18682", "title": "RTPO: Reverse-Turn Policy Optimization for Stabilizing Agentic RL Training", "published": "2026-08-19T08:33:01Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agentic-rl", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B10", "plan_reason": "Agentic RL 与长时序 credit assignment — completed" }, { "arxiv_id": "2608.18681", "title": "Learning What to Fail On: Failure-Mode Contextual Bandits for Adversarial Data Curation", "published": "2026-08-19T08:31:15Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-agent", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18677", "title": "Sanyu Studio: A Multi-Agent System for Art-Historical Narrative Construction", "published": "2026-08-19T08:28:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18675", "title": "An Empirical Benchmark of Deep Time-Series Models for Smart Meter Energy Forecasting", "published": "2026-08-19T08:27:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18671", "title": "Vision-Language Models for Egocentric Video: From Hand-Object Interaction to Embodied AI", "published": "2026-08-19T08:21:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18656", "title": "FlashAttention for Scalable Vector Architectures", "published": "2026-08-19T08:02:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "long-context", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.18645", "title": "Code Health in LLM-Based Test Generation: Effectiveness and Token Efficiency", "published": "2026-08-19T07:47:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18637", "title": "PILOT Technical Report", "published": "2026-08-19T07:35:10Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-taobao", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.18628", "title": "When Safety Overrides Vision: Exploring Dynamics between Vision Influence and Safety Alignment in Vision-Language Models", "published": "2026-08-19T07:22:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "multimodal-llm", "vision-language" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.18613", "title": "CTIFoundry: An Agent-Native Corpus Scaffold for Cyber Threat Intelligence", "published": "2026-08-19T07:00:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18602", "title": "Teach a Molmo2Fish: Towards interactive fish tracking with natural language guidance", "published": "2026-08-19T06:44:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18597", "title": "Off-Manifold Collapse in Guided Protein Language Models", "published": "2026-08-19T06:40:05Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18592", "title": "Infrared Universality of Collective Dynamics across Transformer and State-Space Architectures", "published": "2026-08-19T06:34:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18591", "title": "Can a Lightweight Multimodal Model Estimate LLM Reasoning Performance? A Study for Compute-Optimal Document Inference", "published": "2026-08-19T06:33:33Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18588", "title": "AppEval: A Unified Benchmark for LLM-Based Mobile Application Repair in ArkTS, Swift, and Kotlin", "published": "2026-08-19T06:30:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18586", "title": "OmniHandwritingOCR: A Diagnostic Benchmark for Evaluating Multimodal LLMs in Handwritten OCR Scenarios", "published": "2026-08-19T06:27:05Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18581", "title": "From Storage to Access: Verifiable Activation of Parametric Knowledge in LLMs via Explicit Priming and Implicit Reasoning", "published": "2026-08-19T06:20:34Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18579", "title": "MR-IQA-2: Faithful Image Quality Reflection via Fine-Grained Credit Assignment", "published": "2026-08-19T06:17:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18578", "title": "Compress and Forget: bitsandbytes Quantization Amplifies Proactive Interference in LLMs", "published": "2026-08-19T06:17:13Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "long-context", "model-compression", "post-training", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.18575", "title": "Beyond LLM-Based Reasoning: Lightweight GNNs for Agent Failure Attribution", "published": "2026-08-19T06:16:09Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "llm-architecture", "long-context", "multi-agent", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18565", "title": "SemaPLC: A Project-Grounded, Verification-Gated Agent Harness for PLC Code Generation", "published": "2026-08-19T05:44:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18554", "title": "CentaurBench: Benchmarking LLM Capabilities on Augmenting vs. Automating Real-World Work Tasks", "published": "2026-08-19T05:22:20Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20421", "title": "Six misconceptions about large language models: A minimal model and diagnostic taxonomy", "published": "2026-08-19T04:59:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18539", "title": "Evaluating and Explaining Prompt Sensitivity of LLMs Using Interactions", "published": "2026-08-19T04:45:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18524", "title": "DART-SD: Diamond-topology Aware Retrieval and Tuning for Self-Distillation of Multi-Turn Tool-Calling Agents", "published": "2026-08-19T04:19:21Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-agent", "multi-agent", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18503", "title": "LLM-Powered Predictive Decision-Making for Sustainable Data Center Operations", "published": "2026-08-19T03:50:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18486", "title": "WhiteMatter: All-to-All Cross-Layer Connections via KV Mixing", "published": "2026-08-19T03:24:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18482", "title": "Coverage-Driven RTL Assertion Generation with Formal Exploration and Neuro-Symbolic Refinement", "published": "2026-08-19T03:14:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18479", "title": "COSTA: A Cluster-Centric Paradigm for Annotation-Free Open-Set Semantic Segmentation of Aerial Point Clouds with Domain Shifts", "published": "2026-08-19T03:10:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18474", "title": "OmniAlign: A Unified Multilingual Aligner for Word and Sentence Alignment", "published": "2026-08-19T03:02:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18469", "title": "ERASE: EaRly bAckpropagation SchEdule for Faster Training of Modern Recommendation Systems", "published": "2026-08-19T02:51:08Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18450", "title": "Adaptive Multi-Agent Feature Selection for Personalized Fall Risk Prevention", "published": "2026-08-19T02:25:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18423", "title": "FM-Bench: A Benchmark for Long-Horizon Management with Competing Agents", "published": "2026-08-19T01:33:36Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agent-planning", "llm-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.18412", "title": "JSL-DC: A Word-Level Japanese Sign Language Dataset with Linguist-Derived Descriptions for Distinguishing Confusable Signs", "published": "2026-08-19T00:43:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18410", "title": "Role-Conditioned Sub-Token Routing for Efficient Vision-Language-Action Policies", "published": "2026-08-19T00:38:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18401", "title": "Multimodal Rapport Estimation in Real-World HRI", "published": "2026-08-19T00:15:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18398", "title": "LEDGER: Claim-to-Evidence Trace Graphs for Auditing LLM Agents", "published": "2026-08-19T00:10:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18389", "title": "A Jagged Frontier: Evaluating Robustness of Code Agents to Semantics-Preserving Transformations", "published": "2026-08-18T23:46:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18386", "title": "TTSD-FAR: Test-Time Self-Distillation with Fisher-Anchored Restoration for Missing-Modality Emotion Recognition in LVLMs", "published": "2026-08-18T23:40:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18361", "title": "Figurative and Cultural Knowledge in LLMs: Investigating Cross-Domain Transfer through Fine-Tuning", "published": "2026-08-18T22:25:17Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18360", "title": "One Gate Is Not Enough: Composing Stateful Pre-Action Controls for Agentic AI", "published": "2026-08-18T22:24:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18351", "title": "Task-Conditioned Least-Privilege Learning for Executable Terminal and MCP Agents", "published": "2026-08-18T22:07:23Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18339", "title": "From Inference to Adaptation: A Unified Optimal Transport View of Vision Language Model", "published": "2026-08-18T21:44:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18322", "title": "Multimedia Asset Personalization via Multimodal Embeddings at Netflix", "published": "2026-08-18T21:15:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-netflix", "production-evidence", "recsys-general", "search-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B01", "plan_reason": "8 月工业生成推荐与多模态 — completed" }, { "arxiv_id": "2608.18312", "title": "Artifact-centered Claim-aware Observability for Autonomous Scientific Agents", "published": "2026-08-18T20:47:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18309", "title": "XRF-to-Optical Field-of-View Localization with Vision Language Models", "published": "2026-08-18T20:41:07Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18307", "title": "ComponentBench: Diagnosing Component-Level Failures in Computer-Use Agents", "published": "2026-08-18T20:38:26Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18292", "title": "GuideFetch: A Task Coordination Framework for Concurrent Navigation and Object Retrieval in Assistive Robot Dogs", "published": "2026-08-18T20:16:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20418", "title": "Rigorous Evaluation of Large Language Models for Malaria Drug Discovery: Trade-offs in Performance, Scale, and Resource Utility", "published": "2026-08-18T20:12:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18280", "title": "What Makes Software Issue Resolution Tasks Difficult for Agents?", "published": "2026-08-18T19:59:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18279", "title": "A Comprehensive Review of Large Language Models for Nanophotonics: From Surrogate Modeling to Autonomous Design", "published": "2026-08-18T19:57:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18272", "title": "SeisEvo: Evolution of Seismic Data Reconstruction Algorithms by Agents", "published": "2026-08-18T19:45:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18263", "title": "SIGMA: Symmetry-aware, Intelligent, Geometric, Multi-objective Adaptive Control for Robust, Dependable Traffic Management", "published": "2026-08-18T19:25:52Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18242", "title": "ClosureBench: A Constructive Benchmark for Compositional Graph Reasoning", "published": "2026-08-18T18:36:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18230", "title": "Allocating Recurrent Compute in Looped Language Models", "published": "2026-08-18T18:18:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18072", "title": "Multi-Agent AI System for Radiology Report Structuring and Quality Assurance with Independent Radiologist Evaluation", "published": "2026-08-18T17:57:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18062", "title": "TokEval: A Tokenizer Evaluation Suite", "published": "2026-08-18T17:52:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18058", "title": "Delegation Asymmetry in Agentic Recommender Systems: Measuring Two-Sided Receptivity in Online Dating", "published": "2026-08-18T17:51:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18035", "title": "Plug-and-Play Traffic Element Awareness for End-to-End Autonomous Driving", "published": "2026-08-18T17:29:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18025", "title": "Why GPT-Style Models Do Not Directly Transfer to Symbolic Music: Compression in the Wrong Coordinate System", "published": "2026-08-18T17:20:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18011", "title": "The IOL-AI Challenge: An Open Challenge towards Advancing Linguistic Reasoning", "published": "2026-08-18T17:00:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18009", "title": "Memory Tree Guided Key Frame Querying for Efficient 3D Question Answering", "published": "2026-08-18T16:56:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18008", "title": "Policy-Invariant Reward Shaping from LLM Feedback: A Framework for Hybrid RL Agents", "published": "2026-08-18T16:55:46Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-agent", "llm-architecture", "llm-rl", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.17981", "title": "Recirculation", "published": "2026-08-18T16:30:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17975", "title": "aDSL: Agentic 3D Creation via Joint Agent-Program Design", "published": "2026-08-18T16:27:44Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17950", "title": "Do Large Language Models Play Six Degrees of Separation? Measuring Topological Compression in Long-Context Manifolds", "published": "2026-08-18T16:05:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "long-context", "model-compression", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.17948", "title": "SIGMA: SHAP-Guided Implicit-Trajectory Generation for Metadata-Free LLM-Based AutoFE", "published": "2026-08-18T16:04:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17941", "title": "Efficient RLVR Scheduling via Graph-Structured Online Difficulty Estimation", "published": "2026-08-18T16:01:00Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17933", "title": "EvoTS-Agent: A Self-Evolving LLM Agent for Financial Time Series Change Point Detection", "published": "2026-08-18T15:55:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17926", "title": "PerFact: Perception-Derived Fact Prompting for 3D Brain MRI Report Generation", "published": "2026-08-18T15:47:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17911", "title": "CABLE: Extending the Reach of Memory Retrieval via Complementary Antecedent-Based Linking and Expansion", "published": "2026-08-18T15:40:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17895", "title": "BEAR-Bench: A Bilingual Enterprise and Academic Reasoning Benchmark for Multimodal Models", "published": "2026-08-18T15:29:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17866", "title": "BayesPrompt: human readable prompts that make sense", "published": "2026-08-18T14:59:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17834", "title": "AdaLens: Interactive Storyline for Monitoring and Steering Long-Running Agentic Data Analysis", "published": "2026-08-18T14:34:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17804", "title": "An Empirical Study of Reward Specification and Benchmark Reliability in GRPO-based LLM Unlearning", "published": "2026-08-18T14:04:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17800", "title": "StartupBench: Benchmarking General-Purpose Agents on Market-Validated End-to-End Workflows", "published": "2026-08-18T14:01:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17795", "title": "TraceSQL: Traceable Answerability Estimation for Reference-Free Text-to-SQL Verification", "published": "2026-08-18T13:56:29Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17776", "title": "Debate Training Reduces Reward Hacking in RLAIF", "published": "2026-08-18T13:40:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17756", "title": "D$^2$ACCI: A Dual-Loop Diagnostic Protocol for Evidence-Preserving Agent Memory", "published": "2026-08-18T13:18:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17744", "title": "Thinking in a Low-Resource Language: What SFT Builds, What RL Fixes, What Accuracy Cannot See", "published": "2026-08-18T13:09:03Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17723", "title": "Vision-Language Models for Analog Gauge Reading: An Empirical Study of Specialization, Transfer and Reliability", "published": "2026-08-18T12:47:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17722", "title": "MemCatalyst: Amplifying Data Auditing on Vision-Language Models via Data Poisoning", "published": "2026-08-18T12:46:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17719", "title": "What Aggregate Scores Miss: Measuring Item-Level Regressions in Commercial LLM API Migrations", "published": "2026-08-18T12:44:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17718", "title": "Beyond Suspicious Steps: Ontological Trust in Long-Horizon Agents", "published": "2026-08-18T12:44:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17694", "title": "GADR: Gathering Architecture Decision Records from Meeting Transcriptions", "published": "2026-08-18T12:11:42Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17687", "title": "Mixture-of-Expert Blocks Contain Strong Hallucination Detection Signals", "published": "2026-08-18T12:00:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17671", "title": "Benchmarking Automated Security Patch Backporting: How Far Are We?", "published": "2026-08-18T11:43:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17665", "title": "GraphWake: Group Polarization via Memory-Mediated Polarization Cascade in LLM-Agent Communities", "published": "2026-08-18T11:38:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17659", "title": "MobileWorldSafety: Benchmarking GUI Agent Safety Against Environmental Injection Attacks in Android Apps", "published": "2026-08-18T11:33:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17644", "title": "LLM-Derived Preference Judgments Are Not Self-Consistent", "published": "2026-08-18T11:01:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17618", "title": "From Student Risk Prediction to SC2R: Semantics-Constrained Counterfactual Recourse for Educational Decision Support", "published": "2026-08-18T10:33:21Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17613", "title": "Once Generated, Ranked: End-to-End Generative Slate Recommendation with Unified Semantic-Collaborative IDs", "published": "2026-08-18T10:24:09Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "generative-recommendation", "industrial-ranking", "priority-org-kuaishou", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B01", "plan_reason": "8 月工业生成推荐与多模态 — completed" }, { "arxiv_id": "2608.17605", "title": "Multi-turn Conversational AI from Text to Multimodal Interaction: Data, Models, Evaluation, and Open Challenges", "published": "2026-08-18T10:14:21Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17597", "title": "HarnessRisk: A Lifecycle-Oriented Benchmark for Agent Harness Safety", "published": "2026-08-18T10:03:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17596", "title": "tinyDSM: A Framework for Skill Modeling and Development for Resource-Constrained Millirobots", "published": "2026-08-18T10:03:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17588", "title": "TRUSS: Towards Task-Reliable and User-Safe Automated Agent Skill Generation", "published": "2026-08-18T09:52:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17587", "title": "Write, Execute, Refine: From Skill Followers to Skill Optimizers via Reinforcement Learning from Execution Feedback", "published": "2026-08-18T09:52:48Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17583", "title": "Auditing Exposure to Harmful Content on TikTok using Multimodal Language Models: A Cross-National, Age-Stratified Study", "published": "2026-08-18T09:49:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17574", "title": "Quantifying Risk Under Evolving Uncertainty: Belief-Dependent Robustness for Safe Sequential Decision Making", "published": "2026-08-18T09:36:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17550", "title": "Code as Representation: A Compilable Parsing Paradigm for Academic Documents", "published": "2026-08-18T09:10:43Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17535", "title": "GroupForward: Building Referable 3D Scenes via Instance-Grouped Feed-Forward Gaussian Splatting", "published": "2026-08-18T08:55:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17534", "title": "ArborMem: Navigating Interaction States with Memory Forests", "published": "2026-08-18T08:54:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17528", "title": "Agent Lightning v1.0: Towards Harnessed Agentic RL", "published": "2026-08-18T08:50:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17524", "title": "Evaluating RL Explainability Methods by How Much They Help Fix Bugs in Agents", "published": "2026-08-18T08:46:26Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "process-reward", "reward-model", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.17521", "title": "BrainNorm: A Foundation Model that knows Normal via Semantic Atlas Pretraining", "published": "2026-08-18T08:45:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17514", "title": "SE-MoLoRA: Shared-Expert LoRA Adapters for Domain-Specific Photographic Assessment", "published": "2026-08-18T08:40:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17504", "title": "Agentic Porting, Construction and Initial Verification and Validation of Libraries within the Open Source Unified TRAnsient Multi-Phase Advanced Reactor simulation Kit (Outram Park) Part I: Thermal Hydraulics", "published": "2026-08-18T08:30:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17501", "title": "SGHA: Evidence-Grounded Research Problem Discovery with Local Language Models", "published": "2026-08-18T08:27:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17499", "title": "Towards Better Agents for Multi-Turn User Interaction: The Next User Turn Is More Than Context", "published": "2026-08-18T08:25:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17471", "title": "When AI Designs AI: Innovation or Imitation?", "published": "2026-08-18T07:57:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18183", "title": "Accelerating Visual On-Policy Distillation with Batched Speculative Jacobi Rollouts", "published": "2026-08-18T07:16:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19263", "title": "The Evaluation Context Protocol (ECP): A Portable Contract for AI Agent Evaluation", "published": "2026-08-18T07:14:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17433", "title": "Task-Aware Harness Provisioning for LLM Agents in Mission-Critical Infrastructure Operations", "published": "2026-08-18T07:03:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17427", "title": "Counterfactual Anatomy-guided Spatial-Temporal Decoding for Annotation-Free Hallucination Mitigation in Medical VLMs", "published": "2026-08-18T06:45:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17426", "title": "SemComp-Bench: Benchmarking Semantic Task Completion in Video Generation", "published": "2026-08-18T06:45:07Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17421", "title": "TEAMS: Text-prompted spatiotEmporal dual-heAd Mamba Snake", "published": "2026-08-18T06:43:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17411", "title": "GUPO: Gradient Uncertainty-aware Policy Optimization for Post-Training Large Language Models", "published": "2026-08-18T06:22:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17402", "title": "MoE-ViE: Mixture of Experts Vision Encoder for Efficient Image and Video Understanding", "published": "2026-08-18T05:58:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17393", "title": "LEGO-RL: Harness-Native Reinforcement Learning for Coding Agents", "published": "2026-08-18T05:34:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17379", "title": "PTXBench: Benchmark and Adapt LLMs for GPU Kernel Optimization with Architecture-specific PTX", "published": "2026-08-18T05:14:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17373", "title": "Integrating Novelty and Surprise for Experience Prioritization and Exploration in Image-Based Reinforcement Learning", "published": "2026-08-18T05:07:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17351", "title": "Primitive-Driven Compositional Forensic Visual Prompting for Open-World Face Anti-Spoofing", "published": "2026-08-18T04:18:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17347", "title": "Repetition as Reinforcement: Enhancing Sample Efficiency via Instant Episode Repetition in Reinforcement Learning", "published": "2026-08-18T04:11:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17328", "title": "MS-MFAD : Multimodal large language models for Face Anti-spoofing Detection", "published": "2026-08-18T03:38:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17318", "title": "If, Then, Otherwise: Diagnosing Conditional Branching in Vision-Language Navigation", "published": "2026-08-18T03:22:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17316", "title": "Empowering Compact LLMs with Fusion of Layer-wise Exits for Recommendation", "published": "2026-08-18T03:14:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.17310", "title": "Agentic ESOpt: Fine-Tuning Long-Horizon LLM Agents with Minimal GPU Requirements", "published": "2026-08-18T03:03:53Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17306", "title": "Learning What Not to Learn: Adversarial Disentangled Prompt Tuning for Robust Vision-Language Models", "published": "2026-08-18T03:00:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17289", "title": "PlanPO: Group Planning-Aware Policy Optimization for Multi-Turn Agentic LLMs", "published": "2026-08-18T02:39:53Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-planning", "llm-agent", "multi-agent" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B10", "plan_reason": "Agentic RL 与长时序 credit assignment — completed" }, { "arxiv_id": "2608.17288", "title": "Q-Interference: Memory-Efficient Phase-Aware Quantum-Inspired Attention", "published": "2026-08-18T02:38:17Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-architecture", "llm-rl", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17279", "title": "Key-Frame Reasoning with SAM3: Third Place Solution for the MeViS-Text Track of the 8th LSVOS Challenge", "published": "2026-08-18T02:11:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17268", "title": "Understanding Curriculum Learning in Large Language Models via Cross-Difficulty Optimization Dynamics", "published": "2026-08-18T01:51:20Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17253", "title": "Co-RL: Unsupervised Reasoning Emerges from Diverse Cohort in Multi-agent RL", "published": "2026-08-18T01:16:02Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-agent", "llm-rl", "multimodal-llm", "reward-model", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17237", "title": "Structural Plan-to-Model Conversion with Deterministic Geometry and Guarded Agentic Vision-Language Refinement", "published": "2026-08-18T00:53:31Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17223", "title": "Temporal Leakage in Financial News NLP: A Multi-Architecture Audit with a Regime-Specific M&A Signal", "published": "2026-08-18T00:23:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17220", "title": "PACE: Policy-Attested Contract Execution for Safe AI Agents in Decentralized Finance", "published": "2026-08-18T00:22:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17209", "title": "Teach and Grow: An Agent-Centered Architecture for General Robot Learning", "published": "2026-08-17T23:45:21Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-architecture", "multimodal-llm", "vision-language" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.17205", "title": "Which Source Wins? Task-Dependent Reliance in Vision-Language Models", "published": "2026-08-17T23:38:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17203", "title": "Expressivity In Multimodal Contrastive Learning", "published": "2026-08-17T23:29:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17195", "title": "Graphectory Viewer: A Tool for Process-Centric Analysis of Agentic Software Trajectories", "published": "2026-08-17T23:17:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17188", "title": "Token Optimization and Context Window Management in Multi-Agent AI Workflows", "published": "2026-08-17T22:56:50Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "llm-architecture", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17184", "title": "AISA: AI Safety Assistant Framework for Continuous Improvement of Highway Construction", "published": "2026-08-17T22:48:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17177", "title": "Grounding AI Agents in Contracts: An Empirical Evaluation of Spec-Driven Test Generation", "published": "2026-08-17T22:36:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17170", "title": "Synthesizing Feature Extractors: An Agentic Approach for Algorithm Selection", "published": "2026-08-17T22:12:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17163", "title": "Q-Learning With World Models", "published": "2026-08-17T22:00:42Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17162", "title": "OraclePhys: A Systematic Framework for LLM Fine-Tuning on Structural Mechanics", "published": "2026-08-17T21:56:21Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17159", "title": "A Multi-Surface Consistency Audit of Software Citation Metadata", "published": "2026-08-17T21:52:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17153", "title": "Towards Safer RAG: Only Agents Capable of System 2 Thinking may Access Untrusted Documents", "published": "2026-08-17T21:37:20Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17148", "title": "Authorization Before Context: A Model-Neutral Audience Boundary Against Cross-Audience Memory Leakage in Agentic Systems", "published": "2026-08-17T21:31:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17129", "title": "PROBE: Manipulation-Grounded Visual Question Answering with VLM Agents", "published": "2026-08-17T21:03:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17102", "title": "Emotion Across Speech and Faces: Shared Affective Mechanisms in Multimodal Foundation Models", "published": "2026-08-17T20:26:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17095", "title": "Inference-Time Attention Steering for Vision-Language-Action Driving Models", "published": "2026-08-17T20:03:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17091", "title": "Deep Learning for Cross-Border Electricity Price Forecasting: A Comparative Study", "published": "2026-08-17T19:57:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17084", "title": "Uncertainty-Aware Decision Making in Multimodal Large Language Models", "published": "2026-08-17T19:43:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17075", "title": "Foundation Agents Meet Agentic Deep Research: Evidence-Grounded Clinical Code Forecasting", "published": "2026-08-17T19:24:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17051", "title": "Institution-Specific LLM Prompting Recovers PHI That De-identification Systems and Their Gold Standards Both Miss", "published": "2026-08-17T18:56:04Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "llm-architecture", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17029", "title": "LadderTeam: Dual-Agent Laddering Elicitation Framework", "published": "2026-08-17T18:27:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18171", "title": "Looped Language Models Improve Compositional Tool Calling", "published": "2026-08-17T18:17:54Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-architecture", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.17018", "title": "ORCA: Observability-Grounded Program Repair for Microservice Incidents", "published": "2026-08-17T18:11:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17009", "title": "PowderLine: a programmatic powder diffraction analysis application", "published": "2026-08-17T18:05:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.17007", "title": "SkillEffect: Checked Lowering for Memory-Bounded Agent Tools", "published": "2026-08-17T18:03:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16889", "title": "Don't Drop the BATON: Long-Horizon Robot Manipulation via Agentic Subtask Exploration and Transition-aware Memory", "published": "2026-08-17T17:59:57Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning", "llm-agent", "vision-language" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.16868", "title": "Towards Computational Provenance: Carrying Causal-State Evidence in Generated Text", "published": "2026-08-17T17:50:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16844", "title": "Proteus: Incremental Memory Activation for Long-Context Sequence Modeling", "published": "2026-08-17T17:30:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "long-context", "model-compression" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.16837", "title": "HAF: Adapting Generalist VLAs to Humanoid Whole-Body Loco-manipulation via Hierarchical Action Flow and Spectral Latent RL", "published": "2026-08-17T17:22:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16831", "title": "Policy Iteration with Human Feedback: Bringing Post-Training RL to In-context Learning", "published": "2026-08-17T17:16:43Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16806", "title": "Breaking Planner Integrity Boundary: Enviroment State-Text Injection Attack on LLM-Driven Embodied Agents", "published": "2026-08-17T17:02:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16805", "title": "Diagnosing Dense Same-Class Attribute Misbinding in Large Vision-Language Models", "published": "2026-08-17T17:01:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16801", "title": "When Agents Coordinate: Measuring Coordination in Multi-Agent AI Coding", "published": "2026-08-17T16:57:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16798", "title": "ClawGym II: Exploring Black-Box RL on Agent Harness", "published": "2026-08-17T16:53:03Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16797", "title": "UniDot: A Unified Network for Sequence Modeling and Feature Interaction in Large-scale Recommendation", "published": "2026-08-17T16:52:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16794", "title": "Neurosymbolic Embodied Agents", "published": "2026-08-17T16:50:59Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16775", "title": "Topological Attribution Distance (TAD): Revealing Segment-Level RAG Influence on LLM Output Geometry for Incident Log Analysis", "published": "2026-08-17T16:21:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16747", "title": "Would this change your answer? Evaluating Explanations of LLM Behavior In The Wild with Counterfactual Experiments", "published": "2026-08-17T15:57:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16742", "title": "TDD-Agent: Test-Driven Reasoning for Code Generation", "published": "2026-08-17T15:52:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16739", "title": "Le Critique: Privileged Value Functions for LLM Reinforcement Learning", "published": "2026-08-17T15:49:36Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16733", "title": "GoalEvolve: From Handcrafted Algorithm Priors to Goal-Driven Evolution of Physical Design Algorithms", "published": "2026-08-17T15:45:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16710", "title": "The Ethical Decision Head: Operationalizing Normative Ethics in Autonomous Vehicles via Reinforcement Learning from Human Feedback", "published": "2026-08-17T15:26:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16709", "title": "MIRROR: Multimodal Intelligent Radiology Reasoning and Observation Reporter", "published": "2026-08-17T15:25:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16707", "title": "Semantic Bandits: In-Context Exploration-Exploitation is Biased by Semantic Priors", "published": "2026-08-17T15:25:18Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-agent", "multi-agent", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16704", "title": "Unbiased Recommender Systems with Implicit Feedback", "published": "2026-08-17T15:23:52Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16690", "title": "AnchorScore: A CLIP-Based Diagnostic of MLLM Annotation Difficulty", "published": "2026-08-17T15:10:33Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16686", "title": "Closing the Affective Loop: Multimodal Speaker-Listener Emotion-Dynamics-Aware Empathetic Social Robots", "published": "2026-08-17T15:08:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16671", "title": "Does the LM Head Create a Harmful Gradient Bottleneck? A Causal Test", "published": "2026-08-17T14:59:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16666", "title": "Chronocooked: A Benchmark for Implicit Interval Timing in Reinforcement Learning Agents", "published": "2026-08-17T14:55:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16661", "title": "Turning spectra into images improves plant trait retrieval with 2D-CNNs", "published": "2026-08-17T14:53:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16647", "title": "Every Coin Has Two Sides: On the Dual Nature of Generalization in On-Policy Distillation of Large Language Models", "published": "2026-08-17T14:46:53Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16645", "title": "Reconstruction: A Blind Benchmark for Recovering Research Ideas from Pre-Publication Bibliographies", "published": "2026-08-17T14:44:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16974", "title": "Position: Fairness Failure in Generative Models is an Evaluation Problem", "published": "2026-08-17T14:41:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16637", "title": "PDDLCoder: Agentic PDDL Generation for LLM-Assisted Symbolic Planning", "published": "2026-08-17T14:34:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16630", "title": "The Working Set of a Coding Agent: Coherence Debt in Repository-Scale Tasks", "published": "2026-08-17T14:30:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16622", "title": "HarmTrace: Anchor-Calibrated Decoupled Optimization for Fine-Grained Target Identification in Harmful Memes", "published": "2026-08-17T14:21:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16620", "title": "Palmyra x6 Technical Report: An Agentic, Tool-Use Model Post-Trained via Anchored Supervised Fine-Tuning", "published": "2026-08-17T14:21:03Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16554", "title": "Ask, Condition or Abstain: Reinforcement Learning for Missing-Premise Reasoning", "published": "2026-08-17T13:24:41Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16539", "title": "Listen, Reason, and Segment: Aligning LALMs with Editorial Judgment for Media Chapterization", "published": "2026-08-17T13:14:01Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16536", "title": "DSPrompt: Dynamic Soft Prompt Defense Against M-RAG Corruption", "published": "2026-08-17T13:11:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16514", "title": "Matched Outcomes, Divergent Gaze: How Foveated MLLMs Search Compared to Humans", "published": "2026-08-17T12:54:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "multimodal-llm", "vision-language" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.16513", "title": "MLLM-Guided Semantic Correction for Text-to-Video Generation", "published": "2026-08-17T12:54:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16508", "title": "LLMs for Zero-Shot Threat Detection via Structured Risk Indicators", "published": "2026-08-17T12:47:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16484", "title": "Remote-Sensing City Layout Extraction with MLLM", "published": "2026-08-17T12:23:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16480", "title": "RISE: Roadside Infrastructure Sequence Understanding across 3D Tracking and Structured Vision-Language Reasoning", "published": "2026-08-17T12:22:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16447", "title": "HaReCAP: Habitual-action Grounding for Recursive Large Language Model Agents", "published": "2026-08-17T11:47:03Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-planning", "llm-agent", "multi-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.16419", "title": "PertMind: Eliciting Emergent Biological Reasoning in LLM via Reinforcement Learning on Cellular Perturbation Data", "published": "2026-08-17T11:17:26Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "llm-rl", "post-training", "pretraining-data", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.16411", "title": "Towards Risk-free AI Agent Deployment", "published": "2026-08-17T11:07:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16410", "title": "TRACE-CASH: Trial-History-Conditioned Reinforcement Learning for Adaptive Configuration Exploration in Time-Series CASH", "published": "2026-08-17T11:06:16Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-architecture", "llm-rl", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16407", "title": "POI Recommendation with LLM-Augmented Multi-Graph Learning and Contrastive Alignment", "published": "2026-08-17T11:01:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16402", "title": "A Policy Algebra for Trust-Preserving Agentic AI Execution", "published": "2026-08-17T10:59:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16391", "title": "Ventor-QTest: Threat-Model-Driven Verification of Vendor-Hosted LLM APIs", "published": "2026-08-17T10:41:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16386", "title": "Mint-Agent: Introducing Finance-Native Agentic Foundation Models", "published": "2026-08-17T10:38:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16357", "title": "MELD: A Protocol for Merging Knowledge Across Distributed Agentic Memories", "published": "2026-08-17T10:09:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16349", "title": "AeroCopilotBench: A Two-Tier Benchmark for Evaluating LLM Agents as Aviation Copilots in an Interactive Virtual Cockpit Environment", "published": "2026-08-17T09:53:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16347", "title": "Architecture-Dependent Causal Transfer of Activation States Across Large Language Models", "published": "2026-08-17T09:53:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16344", "title": "IndicQE-APE: A Benchmark for Quality Estimation and Automatic Post-Editing for Indic Languages", "published": "2026-08-17T09:50:52Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "model-compression", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16336", "title": "Beyond Binary Priorities: Multi-Tier SLA Scheduling for Large Language Model Serving", "published": "2026-08-17T09:43:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16333", "title": "Step-Level On-Policy Distillation: Interpolating Between On-Policy Distillation and Supervised Fine-Tuning", "published": "2026-08-17T09:37:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16332", "title": "Unlocking Motion in Expressions: Temporal Calibration for Referring Video Object Segmentation", "published": "2026-08-17T09:36:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16316", "title": "Deep Thought Alignment: Trajectory-Level Latent Distillation for Video Reasoning", "published": "2026-08-17T09:23:49Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "long-context", "multimodal-llm", "on-policy-distillation", "vision-language" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.16302", "title": "Comparing the Quality of Code Generated by Vibe Coding Tools", "published": "2026-08-17T09:14:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16295", "title": "Executable Code Knowledge: Code as a Native, Validation-Carrying Knowledge Representation for AI Coding Agents", "published": "2026-08-17T09:07:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16284", "title": "TransAnyText: Translating Arbitrary Text in E-commerce Images via Structured Visual Generation", "published": "2026-08-17T08:54:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16274", "title": "Decoupled Temporal Encoding for Generative Recommendation", "published": "2026-08-17T08:47:21Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16273", "title": "Foresight-England: Development of a National-Scale Generative AI Model of Electronic Health Records for Medical Event Prediction across the COVID-19 Pandemic", "published": "2026-08-17T08:46:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16263", "title": "Seeing Before Answering: Training-Free Visual Layer Profiling for Vision-Language Models", "published": "2026-08-17T08:35:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16259", "title": "Defake-o3: From Speculative Rationales to Verifiable Evidence for Explainable AIGI Detection", "published": "2026-08-17T08:30:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16251", "title": "SCOUT: Semantic Concept Discovery for Open-Vocabulary Editing of face Recognition Templates", "published": "2026-08-17T08:25:07Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16246", "title": "CompoSkill: Compositional Skill Chain Attacks from Individually Scanner-Passing LLM Agent Skills", "published": "2026-08-17T08:20:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16234", "title": "GaussianDWM++: Language-Grounded 3D Gaussian Driving World Model for Unified Scene Understanding, Editing, and Multi-Modal Generation", "published": "2026-08-17T08:12:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16211", "title": "BaT: Towards Self-Evolving Medical Research Agent with Stage Rubrics", "published": "2026-08-17T07:44:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16207", "title": "Competing at Every Price Point with Agentic Evolution over a Menu of LLMs", "published": "2026-08-17T07:40:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16203", "title": "INSPIRE: A Benchmark for Instruction-Aware Speech Retrieval", "published": "2026-08-17T07:35:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16201", "title": "Multi-Granularity Sentiment Integration for LLM-Based Multimodal Sentiment Analysis", "published": "2026-08-17T07:34:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16185", "title": "LENS: In-Context Search via Latent Evidence Exploration over Dynamic Raw Documents", "published": "2026-08-17T07:04:25Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "agent-memory", "llm-agent", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16181", "title": "MUSE: An Interactive Meta-Agent for Understanding and Steering LLM-powered Data Science Systems", "published": "2026-08-17T06:54:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16178", "title": "Agent-Native Telemetry: Verifiable State-Delta Evidence for Autonomous Operations", "published": "2026-08-17T06:50:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16177", "title": "Measuring Obedience to Authority Across Large Language Models with the Milgram Paradigm", "published": "2026-08-17T06:48:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16168", "title": "QUMem: Personalized Memory for Query-Conditioned User-State Inference in LLM Agents", "published": "2026-08-17T06:34:24Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning", "llm-agent", "long-context", "multi-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.16156", "title": "TRCA: Transition-wise Rubric Credit Assignment for Long-horizon LLM Agents", "published": "2026-08-17T06:19:12Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agentic-rl", "llm-agent", "multi-agent" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B10", "plan_reason": "Agentic RL 与长时序 credit assignment — completed" }, { "arxiv_id": "2608.16114", "title": "HyperSkill: Self-Evolving LLM Agents via Hypergraph-Structured Skill Memory", "published": "2026-08-17T05:10:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16103", "title": "Beyond Similarity Matching: Structured Reasoning for Open-Vocabulary Referring Segmentation in 3DGS", "published": "2026-08-17T04:48:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16094", "title": "Protein Structure Prediction: From Evolutionary Constraints to Generative Modeling", "published": "2026-08-17T04:33:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16081", "title": "SafeGesture: Evaluating Fine-Grained Hand Gesture Understanding in Vision-Language Models through Scenario-Conditioned Safety Interpretation", "published": "2026-08-17T04:25:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16075", "title": "TRACER: Balancing Stability-Plasticity-Cognitivity Trilemma for LLM Enhanced Continual Recommendation", "published": "2026-08-17T04:12:35Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16074", "title": "US-VLA: An Ultrasound Vision-Language-Action Model for Embodied Abdomina", "published": "2026-08-17T04:12:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16073", "title": "GOD: Enhancing Generalization via Deep Grafting for Sequential Recommendation", "published": "2026-08-17T04:11:49Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16072", "title": "Learn What's Left, Not What's Mastered: Saturation Aware Advantage Reweighting for Multi-Reward Policy Optimization", "published": "2026-08-17T04:07:50Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B09", "plan_reason": "Rubric、外部 rollout 与多奖励 RL — completed" }, { "arxiv_id": "2608.16071", "title": "Skill2Query: Exploiting Skill Structure to Generate Pseudo-Queries for Agent Skill Retrieval", "published": "2026-08-17T04:04:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16068", "title": "CAPO: Constraint-Aware Prompt Optimization for LLM Agents", "published": "2026-08-17T04:01:04Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "post-training", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16045", "title": "Walk Before You Run: The Importance of Data Exploration for Data Analysis Agents", "published": "2026-08-17T03:19:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16033", "title": "$R^3$-Bench: LLMs Struggle with Resource-Rational Reasoning under Shared Budgets", "published": "2026-08-17T02:59:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16022", "title": "OpenHarmony Bench: Evaluating LLMs and Coding Agents on OpenHarmony App Development", "published": "2026-08-17T02:27:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16011", "title": "ReRef-3D: A Benchmark for Spatial Referring Expression-Guided 3D Scene Rearrangement", "published": "2026-08-17T02:03:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16010", "title": "Breaking the Compression Barrier: Cross-Architecture Compression Boundary Learning via Reverse Regrowth", "published": "2026-08-17T01:59:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16008", "title": "Spatial Temporal Synergy: Balancing Change and Invariance in Text Driven 3D Human Motion Editing", "published": "2026-08-17T01:54:44Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16002", "title": "From Sequence to Structure: Relational Uncertainty Propagation for LLM Agents", "published": "2026-08-17T01:40:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15984", "title": "A Plug-and-Play 2D Motion Interface for Real-World Motion Language Models", "published": "2026-08-17T00:41:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15971", "title": "The Limits of Binding in Dual Encoders", "published": "2026-08-16T23:54:23Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-architecture", "reward-model", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15965", "title": "Beat the Counter First: A Baseline for Temporal-Graph Anomaly Detectors", "published": "2026-08-16T23:39:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15962", "title": "SEER: Long-Context Reasoning via Selective Visual-Text Compression", "published": "2026-08-16T23:30:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "long-context", "model-compression", "vision-language" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.15956", "title": "Navigation-Informed Embeddings: Dense-Retriever Adaptation from Agent Search Traces", "published": "2026-08-16T23:07:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15949", "title": "Ask to Be Sure: Informative Interactions for Confident Multi-Turn LLM Recommendation", "published": "2026-08-16T22:14:12Z", "tracks": [ "agent", "post-training", "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agentic-rl", "generative-recommendation", "llm-agent", "llm-recommendation", "llm-rl", "multi-agent", "preference-optimization", "recsys-general", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.18167", "title": "Adversarial Review: Structured Disagreement for Grounded Agentic Code Review", "published": "2026-08-16T21:58:17Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15931", "title": "PLSQLBench: Benchmarking LLM Systems for Executable Procedural Database Programming", "published": "2026-08-16T21:03:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15888", "title": "Bounded Agents: Delegation Security for Multi-Agent AI Systems", "published": "2026-08-16T18:38:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16959", "title": "MagViT: Interpretable Multi-Magnification Transformers with Patient-Level Model Selection for Breast Histopathology", "published": "2026-08-16T18:20:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15879", "title": "When Less Is Enough: Context Selection and Prompting Strategies for Bengali News Headline Generation", "published": "2026-08-16T18:05:44Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15877", "title": "Dear Algo: A Precision-First Agentic Intent Layer for Unified Search and Recommendation", "published": "2026-08-16T17:59:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15871", "title": "Large Language Models as Implicit Sociological Models: Reconstructing Voting Behaviour from Sociodemographic Profiles", "published": "2026-08-16T17:39:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15869", "title": "Beyond Visual CoT: Internalized Visual Thinking for Proactive Video Reasoning", "published": "2026-08-16T17:35:38Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "post-training", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15868", "title": "CoupVisor: Strategy Optimization by Round and Challenge Decision Support", "published": "2026-08-16T17:29:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15863", "title": "Scaling Manual-Grounded Appliance Manipulation with Data Synthesis and Unified Planning", "published": "2026-08-16T17:15:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15844", "title": "MicroVerse: An Instrument for Measuring Self-Authored Identity Drift in Long-Horizon Multi-Agent Language-Model Simulations", "published": "2026-08-16T16:31:42Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18166", "title": "TractoGraphVLM: A Unified Vision-Language Framework for White Matter Tractography", "published": "2026-08-16T16:14:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15834", "title": "Schema-Agnostic Graph Reasoning Agent for Hybrid Knowledge Graphs", "published": "2026-08-16T16:10:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15797", "title": "KV-Rescue: Recovering Reasoning Language Model KV Eviction Loss via Stepwise Interleaving", "published": "2026-08-16T15:23:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15763", "title": "TaoLive Digital Avatar Agent Technical Report: Training Agents to Evolve with Their Harness", "published": "2026-08-16T14:32:56Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "llm-rl", "on-policy-distillation", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15755", "title": "Intent-Driven Situation Tracking for User-Centric Multi-Turn Agents", "published": "2026-08-16T14:14:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15710", "title": "Beyond Single Object: Learning 3D Relations with Large Language Models", "published": "2026-08-16T12:29:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15703", "title": "HyMem: Hierarchical Context Management for Long-Horizon Agents via Information Isolation", "published": "2026-08-16T12:15:01Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agent-planning", "llm-agent" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B11", "plan_reason": "Agent 记忆、工具规划与自进化系统 — completed" }, { "arxiv_id": "2608.15698", "title": "ConceptFormer: Learning Adaptive Latent Concepts for Query-Document Alignment in Visual Document Retrieval", "published": "2026-08-16T12:07:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15693", "title": "Large Models for Small Devices: Recent Advances and Empirical Analysis of Edge AI Deployment", "published": "2026-08-16T11:52:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15678", "title": "Where Accountability Lives: Mapping Human Responsibility to Workflow Artifacts in Agentic Software Development", "published": "2026-08-16T11:11:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15673", "title": "PL-Guard: Probabilistic Logic Reasoning for LLM Guardrails", "published": "2026-08-16T10:58:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15669", "title": "Large Discovery Models: Empirically-grounded Model-Based Open-Ended Search", "published": "2026-08-16T10:27:06Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-architecture", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15654", "title": "When Stories Evolve: Benchmarking LLM Storytelling Across Agent Architectures in Open-Ended World Simulations", "published": "2026-08-16T09:46:45Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-agent", "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15632", "title": "Sparse Prototype Code Underlies Classification and Prediction Across Modalities", "published": "2026-08-16T08:53:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15630", "title": "Do Assessment Instruments Measure the Same Thing for Humans and LLMs? A Latent Structure Analysis", "published": "2026-08-16T08:52:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15605", "title": "AlloEgo-VLM: Disambiguating Allocentric and Egocentric Reference Frames in Vision-Language Models", "published": "2026-08-16T08:06:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15602", "title": "FluxBin: Flexible LUT-based Ultra-low-bit LLM Inference by Algorithm-Kernel Synergy", "published": "2026-08-16T08:01:06Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-architecture", "model-compression", "post-training", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.15595", "title": "AutoSQL: Extracting SQL Templates from Imperative ORM Code in Large-Scale Repositories", "published": "2026-08-16T07:32:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15591", "title": "Agent Gym: A Framework for Continuous Evaluation and Evolution of LLM Agents Through Human-in-the-Loop Feedback", "published": "2026-08-16T07:27:26Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15580", "title": "From Generalist to Specialist: A Context-Fusion Framework for Endoscopic Polyp Reporting with a Frozen VLM", "published": "2026-08-16T07:11:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15579", "title": "Kozuchi Agent: A Language-Agnostic Open-Weight Agent for Software Repair", "published": "2026-08-16T07:06:37Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-agent", "software-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.15574", "title": "Catching Hallucinated Citations in Video-LLM Question Answering: A Self-Verification Pipeline and Verifier Ablation Study", "published": "2026-08-16T06:42:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15567", "title": "SchurQuant: Groupwise Discrete Optimization for Layer-Wise LLM Quantization", "published": "2026-08-16T06:29:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15565", "title": "Admission Without Answers: Label-Free Certification and Experience Learning for LLM-Based Optimization Modeling", "published": "2026-08-16T06:18:54Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15549", "title": "MistyPilot: Enabling Social-Robot Control through Multi-Agent LLM Skill Orchestration", "published": "2026-08-16T05:45:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15539", "title": "CrossView: Can Vision-Language Models Reason Across Cameras?", "published": "2026-08-16T05:25:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15530", "title": "Why Summaries Turn Neutral: Policy Attribution for Sentiment Drift in Reinforcement Learning from Human Feedback", "published": "2026-08-16T04:56:03Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15517", "title": "GLaQ: Grounding Latent Queries in Visual Evidence for Multimodal Reasoning", "published": "2026-08-16T04:01:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15516", "title": "UniFed-VLM: Federated Instruction Tuning for Vision-Language Models with Multiple Heterogeneity", "published": "2026-08-16T04:01:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15509", "title": "Temporal Logic Guided Universal Task Representations for Reinforcement Learning", "published": "2026-08-16T03:28:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15507", "title": "Do Language Models Consistently Encode the Current Year?", "published": "2026-08-16T03:22:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15491", "title": "HxAgent: Iterative Agent Planning for End-to-End Web Application Testing", "published": "2026-08-16T02:40:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15483", "title": "Measuring Structured Predictability in Neural Training Dynamics: A Cross-Regime Study", "published": "2026-08-16T02:17:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15459", "title": "Not All Attention Is Equal: A Quantitative Survey of the EEI Trade-off", "published": "2026-08-16T00:37:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15456", "title": "AlignJEPA: Predictive Vision-Language Alignment for Remote Sensing Foundation Models", "published": "2026-08-16T00:27:07Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15448", "title": "Language models suffer from a curse of ambiguity", "published": "2026-08-15T23:22:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15432", "title": "Does the Proof Prove It That Way? Faithful Formalization of Elements Proofs", "published": "2026-08-15T22:20:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15429", "title": "SAGA: Structure-Attended Generative Action Embedding Model that encodes Multi-Surface User Action Sequences", "published": "2026-08-15T22:11:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15425", "title": "NumerosityVLM: A Cognitively Inspired Benchmark for Interpreting Numerosity Representations in Vision-Language Models", "published": "2026-08-15T22:01:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15412", "title": "Invariant Pretraining for Robust Code Representations", "published": "2026-08-15T21:01:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15410", "title": "FloodReasonBench: Benchmarking VLM Reasoning Segmentation for Embodied Flood Response at the Edge", "published": "2026-08-15T20:56:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15404", "title": "CBX-Bench: A Human-Aligned MLLM Council for Benchmarking Concept Bottleneck Model Explanations", "published": "2026-08-15T20:21:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20414", "title": "StateSight: Benchmarking Latent Spatial-State Reconstruction in Vision-Language Models", "published": "2026-08-15T20:13:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15391", "title": "TwinGridShield: Consequence-Aware Runtime Authorization for LLM Grid-Agent Actions", "published": "2026-08-15T19:44:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15389", "title": "Agentic-SQL Revisited: Autonomy-Based Taxonomy and Empirical Benchmark Analysis for LLM Text-to-SQL", "published": "2026-08-15T19:40:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15382", "title": "Grounding Healthcare LLMs in a Causal Knowledge Graph: Framework, Metrics, and a Cardiovascular Pilot", "published": "2026-08-15T19:21:51Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15360", "title": "SAPE: Sandwich Adapters for Parameter Efficiency in Large Language Model Fine-Tuning", "published": "2026-08-15T18:32:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15353", "title": "Decomposing Whole Slide Image Report Generation with Graph-Constrained Multiple Instance Learning Workflows", "published": "2026-08-15T18:16:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15304", "title": "Understanding Cognition-Induced Risks in Agentic AI Systems", "published": "2026-08-15T16:16:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15303", "title": "Divergent-Convergent Reasoning: Scaling Test-Time Compute through Structured Solution Synthesis", "published": "2026-08-15T16:13:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15299", "title": "MAPLE: MoE Adaptive Plug-and-play Layer-wise Expert allocation", "published": "2026-08-15T16:10:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15286", "title": "No Task Fails Every Time: Why One-Shot Audits Are Structurally Blind to Agent Damage", "published": "2026-08-15T15:33:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15242", "title": "LongRCA Bench: Diagnosing Responsible Roles and Root Causes in Long-Horizon Agent Failures", "published": "2026-08-15T14:06:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15238", "title": "UC-VLM: Consistency-Driven Learning for AI-Generated Image Detection with Vision-Language Large Models", "published": "2026-08-15T13:53:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15193", "title": "Valhalla: A Layered Knowledge-State and Service-Governance Framework for Long-Term Scientific Knowledge Work", "published": "2026-08-15T12:21:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15175", "title": "LAPF: LLM-Agent-Based Path Finder Using the UAVScenes Dataset", "published": "2026-08-15T11:28:21Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agent-planning", "agentic-rl", "llm-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.15165", "title": "SkillCommit: Evolving Agent Skills through Behaviorally Validated Scope Expansion", "published": "2026-08-15T11:03:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15145", "title": "ACTS-SQL: Agentic and Critic-Oriented Tree-Structured SQL Correctness with Large Language Models", "published": "2026-08-15T09:43:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15129", "title": "Left-Branching Transformers Excel at Right-Branching Languages: Data Shapes Word Order Preferences in Language Models", "published": "2026-08-15T09:03:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15127", "title": "From LLM Inference to Agentic Workloads: Characterization and Implications for Serving Systems", "published": "2026-08-15T09:02:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15117", "title": "Anatomy of a Quantized Agent: VRAM Stability and Forecasting in Code-Synthesis Agentic Workloads", "published": "2026-08-15T08:39:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15115", "title": "Perspective-Invariant Attack with Enhanced Transferability of Adversarial Examples", "published": "2026-08-15T08:32:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15112", "title": "Probability-Preserving Transformer for the Time-Dependent Schrödinger Equation", "published": "2026-08-15T08:29:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15109", "title": "Constraint-Aware Synthetic Tabular Data Generation via Inter-Column Constraint Discovery with LLM Agents", "published": "2026-08-15T08:19:54Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15108", "title": "Beyond Direct Access: Resource Hijacking in LLM Agents", "published": "2026-08-15T08:16:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15085", "title": "Why Vision Fails as a Universal Bridge: Rectifying Modality Asynchrony in Multilingual MLLMs", "published": "2026-08-15T07:04:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15082", "title": "Beyond Thresholds: A Quality-Aware Decision Intelligence Framework for Cold Chain IoT Systems", "published": "2026-08-15T06:59:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15080", "title": "A Pilot Study of Autocompleting Tokenizers", "published": "2026-08-15T06:56:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15075", "title": "SA-GEM: Scale-Adaptive and Geospatial Evidence-Modulated Token Pruning for Efficient Remote Sensing Large Vision-Language Models", "published": "2026-08-15T06:50:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15071", "title": "Evo-Harness: Context-to-Harness Skill Compilation for Self-Evolving Agents", "published": "2026-08-15T06:43:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15061", "title": "Do Visual Grounding Decoders Need Feed-Forward Networks? A Controlled Study over Frozen Vision-Language Features", "published": "2026-08-15T06:21:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15058", "title": "MEDR: Query-Independent Frame Selection via Multi-Signal Event Modeling and Dynamic Rescoring", "published": "2026-08-15T06:10:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15046", "title": "Certifying Compressed Language Models: An Audit and a Statistical Toolkit", "published": "2026-08-15T05:15:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15045", "title": "MOSS-VL Technical Report", "published": "2026-08-15T05:12:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15037", "title": "Prototype-Rectified Iterative Self-supervised Manifold Denoising under Severe Acoustic Shift", "published": "2026-08-15T04:42:14Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15032", "title": "Handoff-H1: An Orchestrated Vision-Agent System for Material Quantity Takeoff from Construction Blueprints", "published": "2026-08-15T04:34:51Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-agent", "llm-architecture", "tool-agent", "vision-language" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.15016", "title": "Hierarchical Agentic Incident Response with Digital-Twin-Validated Attack Inference", "published": "2026-08-15T04:01:27Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-planning", "llm-agent", "multi-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.15012", "title": "SysEvolve: An AI-native, safe, autonomous adversarial attack-defense co-evolutionary system", "published": "2026-08-15T03:47:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15009", "title": "ForceU-VLA: A Force-Aware Vision-Language-Action Model for Embodied Ultrasound Scanning", "published": "2026-08-15T03:39:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15008", "title": "Harness the Memory: A Holistic Evaluation of Memory Substrates in Memory Agents", "published": "2026-08-15T03:29:48Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-agent", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15006", "title": "MetaReason: Precise Interleaved Multimodal Reasoning via Editing Meta Information for Solving Geometry Problems", "published": "2026-08-15T03:24:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15004", "title": "FZ-VLM: A Two Stage Florence-Zephyr Vision Language Model Framework for Pulmonary Nodule Characterization and Clinical Decision Making", "published": "2026-08-15T03:22:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.15002", "title": "NPU Offloading of a Frozen Visual Encoder for Robot Policy Training", "published": "2026-08-15T03:19:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14991", "title": "Risk-Adaptive Edge--Cloud Visual Reasoning for Communication-Efficient Autonomous Driving", "published": "2026-08-15T02:36:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14963", "title": "Command-Space Counterfactual Explanations for Pareto-Conditioned Reinforcement Learning", "published": "2026-08-15T01:32:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14945", "title": "Trust Is Not Enough: Influence Calibration for On-Policy Self-Distillation in Agentic RL", "published": "2026-08-14T23:56:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14944", "title": "SkillComposer: Learning Reusable Skills for Natural-Language Robot Programming", "published": "2026-08-14T23:49:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14927", "title": "LLMs Can Predict Failure Risk, But Struggle to Predict Which Collaboration Protocol Pays Off: Cost-Aware Protocol Routing Across Reasoning Tasks", "published": "2026-08-14T22:35:37Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-agent", "multi-agent", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14924", "title": "PaSTel: Anchoring Histology in Spatial Transcriptomics via Multi-Scale Hierarchical Bio-Prior Contrastive Pretraining", "published": "2026-08-14T22:26:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14905", "title": "How Do Agents Fail on AutoResearch: End-to-End Diagnostic Evaluation on 100 Real-World Frontier Research Tasks", "published": "2026-08-14T21:39:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14886", "title": "Where Does Retrieval Fail? Evaluating RAG Architectures for Agricultural Advisory", "published": "2026-08-14T20:50:24Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14876", "title": "Workspace Topology as an Attack Vector in Agentic Coding Assistants", "published": "2026-08-14T20:30:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14863", "title": "Evaluating Agentic Code Repair Capabilities in Distributed Systems", "published": "2026-08-14T19:59:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14854", "title": "Zero-MELO: Test-Time Evidence Calibration with Multimodal LLMs for Zero-Shot Micro-Gesture Recognition", "published": "2026-08-14T19:50:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14851", "title": "Discovering High-Quality Chess Puzzles with Offline Reinforcement Learning", "published": "2026-08-14T19:46:29Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14838", "title": "The Recall Trap: A Recall-Maximizing Retriever Configuration Reduces Issue Resolution in Fixed-Budget Code Context", "published": "2026-08-14T19:22:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14835", "title": "OvDSGG: End-to-End Open-Vocabulary Dynamic Scene Graph Generation", "published": "2026-08-14T19:12:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14825", "title": "Emergent Misaligned Communication in Long-Horizon Multi-Agent LLM Commerce", "published": "2026-08-14T18:56:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14822", "title": "Imagining Recovery: Inference-Time Counterfactual Realignment for Vision-Language-Action Models", "published": "2026-08-14T18:49:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14813", "title": "Beyond the pale: Assessing prevalence and contents of extremist speech in LLM training data", "published": "2026-08-14T18:36:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14797", "title": "Beyond Tokens: A Survey on Decoding Methods for Large Language and Vision-Language Models", "published": "2026-08-14T18:08:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14783", "title": "MegaParts: Scaling Part-Aware 3D Object Generation to 300 Parts via Token-Efficient Autoregressive Modeling", "published": "2026-08-14T18:00:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14528", "title": "Handover of In-Context Learning State Across Session Boundaries", "published": "2026-08-14T17:47:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14527", "title": "Validating LLM-Modernized Scientific Software Through Differential Fault Injection", "published": "2026-08-14T17:45:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14776", "title": "NRCD: An Open Database of Collegiate Running with Unified Performance Standardization", "published": "2026-08-14T17:36:02Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14509", "title": "Split the Labor: Separating Evidence Interpretation from Decision Aggregation", "published": "2026-08-14T17:24:55Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-architecture", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14505", "title": "RecipeNet: A Hierarchical Transformer for Recipe Data", "published": "2026-08-14T17:18:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14498", "title": "Rollplex: Cross-Phase GPU Spatial Sharing for Vision Language Model Post-Training", "published": "2026-08-14T17:13:34Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14465", "title": "You Only Pass Once: Answering and Abstaining Together in a Single Forward Pass of a Frozen Language Model", "published": "2026-08-14T16:44:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14452", "title": "SheetCompass: Hierarchical Relation Graphs for Agentic Spreadsheet Reasoning", "published": "2026-08-14T16:39:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14443", "title": "Designing Compact Neural Architectures via Neuron Gating and Mixed Activation", "published": "2026-08-14T16:28:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14774", "title": "p-Spin Glass Network Efficient Single-Batch Continual Learning", "published": "2026-08-14T16:24:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14435", "title": "Style or Signature? Artist-Disjoint Evaluation of Style Classification in Frozen Vision Embeddings", "published": "2026-08-14T16:18:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14399", "title": "Whose doctor does the AI recommend? An algorithm audit of reputation and demographic signals in large language model-assisted physician choice", "published": "2026-08-14T15:39:10Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14385", "title": "DeaMoE: Efficient MoE Structure for Fast Small-Batch Decoding", "published": "2026-08-14T15:25:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14380", "title": "AgentRewind: Recoverable Execution for Long-Horizon LLM Agents", "published": "2026-08-14T15:20:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14377", "title": "A Survey of Large Models in Sports", "published": "2026-08-14T15:17:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14354", "title": "ScienceFlow: A long-horizon agent for ML research, scientific discovery and beyond", "published": "2026-08-14T14:54:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14352", "title": "ATLAS: Discovering Agent Strategies through LLM-Guided Abstraction and Automata Learning", "published": "2026-08-14T14:51:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14339", "title": "Clearing the Fog: Towards Installing and Refining Proactive Exploration Capabilities in LLM Agents", "published": "2026-08-14T14:24:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14312", "title": "Envs-FORGE: Frontier-Optimized Reward-Grounded Environment Synthesis for Agent RL", "published": "2026-08-14T13:54:22Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14309", "title": "Spatial Message Passing in Language Space for Pathology Image Interpretation", "published": "2026-08-14T13:48:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14286", "title": "Seeing Red, Thinking Bad: Color Bias in Vision Language Models", "published": "2026-08-14T13:14:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14277", "title": "SimpleOPD: Simple Tokenizer-Agnostic On-Policy Distillation for Long-Context Reasoning", "published": "2026-08-14T12:57:02Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14270", "title": "TimeSage-EV: A Live Benchmark for Agentic Time Series Analysis in Evolving Environments", "published": "2026-08-14T12:53:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14262", "title": "On the Robustness of Temporal Vision-Language Models for Surgical Endoscopy Videos", "published": "2026-08-14T12:37:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14767", "title": "NARRATE: A Multimodal Real-World Australian Driving Dataset for Human-Centred Explanations in Automated Driving", "published": "2026-08-14T12:35:24Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14252", "title": "Grounding Without Corrective Control: Truth-Tracking Profiles for Large Language Models", "published": "2026-08-14T12:30:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14229", "title": "The More Popular, The Harder to Forget: Adaptive Popularity for LLM Unlearning", "published": "2026-08-14T12:03:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14226", "title": "RankT2I: A Submodular Framework for Discovering Interpretable and Diverse Semantics in Text-to-Image Models", "published": "2026-08-14T11:59:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14765", "title": "Agentic Data Cleaning Without a Clean Reference: An Experimental Study of Capabilities and Trade-offs", "published": "2026-08-14T11:55:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14221", "title": "MathForm: Scaling Mathematical Autoformalization with Knowledge Retrieval and Verification-Guided Refinement", "published": "2026-08-14T11:51:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14198", "title": "MINT: A Universal Zero-Shot Predictor for Transaction Data", "published": "2026-08-14T11:17:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "multimodal-llm", "pretraining-data" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.14191", "title": "KV Cache Compression Through the Lens of Transform Coding", "published": "2026-08-14T11:08:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14763", "title": "Cross-Modal Ultrasound-MRI Learning for Fetal Brain Ventricular Volumetry and Abnormality Screening", "published": "2026-08-14T10:33:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14150", "title": "Leading-Silence Augmentation and Multi-Stage Synthetic Supervision for the Second MLC-SLM Challenge", "published": "2026-08-14T10:00:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14135", "title": "AgilePE: Autonomous UAV Pursuit-Evasion via Self-Play Reinforcement Learning", "published": "2026-08-14T09:41:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14131", "title": "LegacyWorld: Atomicity-Aware Evaluation of GUI Agents for Legacy Workflows", "published": "2026-08-14T09:36:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14130", "title": "AlignFace: Human-Aligned Face Similarity Metric with Interpretable Concept Relations", "published": "2026-08-14T09:36:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14122", "title": "Reinforcement Learning-Based Production Scheduling in an Industry-Based Coating Scenario Using the Digital Model Playground", "published": "2026-08-14T09:28:20Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14114", "title": "Learning to Run Power Networks: Effective AlphaZero-inspired Topological Control", "published": "2026-08-14T09:16:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14109", "title": "A Graph-Based Reinforcement Learning Framework for Structured Drift Diagnosis and Recovery in Autonomous LLM Agents", "published": "2026-08-14T09:12:14Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-agent", "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14068", "title": "MACS: A Hybrid Multi-Agent Framework for Reliable Conversational E-Commerce Recommendation", "published": "2026-08-14T08:26:19Z", "tracks": [ "agent", "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-agent", "llm-recommendation", "multi-agent", "recsys-general", "web-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.14066", "title": "TenderKG", "published": "2026-08-14T08:24:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14055", "title": "HERMES: a multi-agent framework for structured knowledge extraction from ultra-long documents in geoscience", "published": "2026-08-14T07:59:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14047", "title": "Evolve Vision-Language-Action Model into an Agent with On-the-fly Tool-use", "published": "2026-08-14T07:53:18Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14757", "title": "KHiM-Mamba: Injecting Pathology Knowledge into Mamba via Hidden-State Modulation for Whole Slide Image Analysis", "published": "2026-08-14T07:45:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14036", "title": "Demystifying Agent Skills: Why They Work-Until They Don't", "published": "2026-08-14T07:26:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14035", "title": "Agent-Orchestration in Autonomous Chip Design", "published": "2026-08-14T07:26:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14029", "title": "S2Dialog: Multimodal Dialogue Retrieval with Semantic and Acoustic-Style Modeling", "published": "2026-08-14T07:19:24Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14024", "title": "SSP: An Event-Matched Syn2Sim2Phy Cross-Domain Evaluation Framework for Autonomous Driving VLA Models", "published": "2026-08-14T07:15:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14016", "title": "Content Based Video Narration of Gameplay with Vision Language Models", "published": "2026-08-14T07:03:30Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14015", "title": "MedClaw: Heuristic Agent Harness for Long-Horizon Surgical Video Reasoning", "published": "2026-08-14T07:03:06Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14011", "title": "EchoRec: Multi-Item Prediction-Empowered Generative Recommendation via Cycle-Consistent Preference Alignment", "published": "2026-08-14T06:59:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13990", "title": "Content Depth Matters in Short-Video Recommendation: Rethinking the Attention Economy", "published": "2026-08-14T06:13:40Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13980", "title": "FIRM: Fine-Grained Intra-Token Representation of Masks for Remote Sensing Reasoning Segmentation", "published": "2026-08-14T05:40:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13974", "title": "ProFocus: Interpreting Affective Experience in Artistic Images with Progressive Visual Focusing", "published": "2026-08-14T05:34:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13973", "title": "Rethinking Auxiliary Modalities in Multimodal Zero-shot Anomaly Detection: From Semantic Fusion to Conditional Modulation", "published": "2026-08-14T05:33:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13969", "title": "PPOM: Marginalizing Patch-Grid Phase for CLIP-Based Generalizable Vision-Language Prompt Tuning", "published": "2026-08-14T05:31:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13966", "title": "QUASAR: Lowering the Loss Floor of Quantization-Aware Training with Loss-Aware Reconstruction", "published": "2026-08-14T05:29:58Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13940", "title": "AI Research Preference Models", "published": "2026-08-14T04:20:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13928", "title": "CoSA: Context-Aware Severity Assessment via Context Analysis with Large Language Models", "published": "2026-08-14T03:57:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13926", "title": "Never the Number: Structural Abstention for AI Systems Whose Answers Are Consumed as Fact", "published": "2026-08-14T03:53:21Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "llm-architecture", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13921", "title": "When Personal Memory Has No Single Answer: Evaluating LLM Agents under Irreducible Conflict", "published": "2026-08-14T03:48:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13900", "title": "Agentic Transaction: Towards ACID-Compliant Agent Systems", "published": "2026-08-14T03:13:54Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "llm-agent", "llm-architecture", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.13889", "title": "Consensus-gated Multi-Agent Neural Architecture Search for Seismic Fault Segmentation", "published": "2026-08-14T02:41:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13884", "title": "Engineering Signals of Human-AI Collaboration in the Agentic Coding Era: A Longitudinal Analysis of 33,228 Pull Requests from vLLM and SGLang with Implications for Biomedical AI Agents and Bioinformatics Pipeline Developmen", "published": "2026-08-14T02:22:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13867", "title": "Engineering Reliable Coding Agents: Evaluating and Operating the System Around the Model", "published": "2026-08-14T01:34:25Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13854", "title": "Bootstrapping Niche Multilingual Code Translation via Reinforcement Learning with Execution-Based Verifiable Supervision", "published": "2026-08-14T00:56:41Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "preference-optimization", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.13833", "title": "AdsWorldEngine: A Self-Evolving Conversational Advertising Agent through Orchestrator and Tool Coevolution", "published": "2026-08-13T23:53:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13831", "title": "VoiceChat-TTS: A Low-Latency Continuous Speech Synthesis Model for Interactive Agents", "published": "2026-08-13T23:37:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13791", "title": "VLM- and LLM-Driven Multi-Agent System for PET Image Denoising", "published": "2026-08-13T21:50:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13787", "title": "From Passive Delegates to Strategic Negotiators: Reinforcing Social Reasoning in Small Language Models with SocialRL", "published": "2026-08-13T21:39:12Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13767", "title": "Simulation-Aware In-Context Policy Improvement for LLM-Aided Analog Layout Refinement", "published": "2026-08-13T20:46:17Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13766", "title": "ChartProbe: A Diagnostic Study on Visual Reasoning through Perception, Grounding, and Simple Reasoning", "published": "2026-08-13T20:44:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13760", "title": "Amplified Does Not Mean Predictive: Reasoning Behaviors in Thinking Models", "published": "2026-08-13T20:37:59Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "reward-model", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13724", "title": "Architecture and Affordances of PLAUD: Performative Latents and Unsupervised DDSP", "published": "2026-08-13T19:36:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13711", "title": "TRUE-Colon: Exposing a Consistent Transfer Asymmetry in Real-Time Polyp Detection", "published": "2026-08-13T19:13:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13702", "title": "SAGE: Surrogate-gradient Adaptation via Attention-Guided Entropy for Spiking Transformers", "published": "2026-08-13T18:51:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13698", "title": "GRPO Beyond English: A Large-Scale Study of GRPO in Non-English and Multilingual Settings", "published": "2026-08-13T18:43:33Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13690", "title": "MedPlex: Deep Vision-Language Co-Adaptation for Clinically Grounded Medical Segmentation", "published": "2026-08-13T18:37:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13681", "title": "Fine-Tuning Qwen3-27B for C-to-Rust Code Translation: A Three-Stage Curriculum of Pretraining, Debugging-Aware SFT, and Task-Specific SFT", "published": "2026-08-13T18:24:02Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "llm-agent", "software-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.13675", "title": "From BERT to Frontier Agents: Eight Years of Language-Model Progress, the Collapse of the Capability-Cost Curve, and the Rise of Task-Targeted Models", "published": "2026-08-13T18:16:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13671", "title": "PROVE: Training-Free Prompt Recovery using Verifiable Evidence", "published": "2026-08-13T18:08:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13667", "title": "Second Thought: Reasoning in Parallel as LLM Agents Act and Observe", "published": "2026-08-13T18:04:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13662", "title": "Ontology-Grounded Project Memory for Coding Agents", "published": "2026-08-13T18:03:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13560", "title": "AutoDesign: Meta-Harness Optimization for Long-Horizon Agentic Design", "published": "2026-08-13T17:59:57Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "multimodal-llm", "preference-optimization", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13547", "title": "QuoteBench: How Matched Scores Can Hide Command-Path Failures", "published": "2026-08-13T17:57:20Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13545", "title": "LittleLearner: Language Models Under Pedagogically Controlled Knowledge Exposure", "published": "2026-08-13T17:56:12Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13522", "title": "Vero: Can AI Agents Build Formally Verified Software Repositories?", "published": "2026-08-13T17:41:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13521", "title": "Exponential quantum advantage for learning signals with a single qubit", "published": "2026-08-13T17:40:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13517", "title": "DFM Mimir v1: An Open HRM Delivering Frontier Performance at 1B Parameters Using Only Permissible Post-Training Data", "published": "2026-08-13T17:37:53Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13515", "title": "Measuring Task-Agnostic Training Data Influence Across Language Model Pretraining", "published": "2026-08-13T17:36:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13513", "title": "TabSOM: A tabular-to-image encoding method based on self-organizing maps", "published": "2026-08-13T17:35:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13505", "title": "Intern-S2-Preview: Scientific Agentic Foundation Model", "published": "2026-08-13T17:31:28Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agentic-rl", "llm-architecture", "llm-rl", "multimodal-llm", "on-policy-distillation", "opd", "post-training", "pretraining-data", "vision-language" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.13489", "title": "DreamX-Phi 1.0: Action-Conditioned Video World Model for Robotic Manipulation", "published": "2026-08-13T17:18:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13482", "title": "Synthetic Persona Pretraining: Alignment from Token Zero", "published": "2026-08-13T17:12:04Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "post-training", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13476", "title": "MARC v1: An Open-Source Multi-Agent Framework for Clinical AI Reasoning and Coordination", "published": "2026-08-13T17:00:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13472", "title": "AaLLM: An End-to-End Analog Circuit Design Framework from Topology Generation to Sizing Using Large Language Models", "published": "2026-08-13T16:57:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13463", "title": "MLLM-Routed Heterogeneous Ensembles for Robust Cross-Dataset Image Classification", "published": "2026-08-13T16:45:24Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "llm-agent", "llm-architecture", "multi-agent", "multimodal-llm", "vision-language" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.13453", "title": "UniTexture: Cross-Task Universal Adversarial Textures for Vision-Language-Action Models", "published": "2026-08-13T16:38:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13441", "title": "Edit2TikZ: A Comprehensive and Challenging Benchmark for Scientific Figure Editing with TikZ", "published": "2026-08-13T16:27:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13428", "title": "RAIL: An Automatic Classifier of the Artificial Intelligence Readiness Level", "published": "2026-08-13T16:17:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13426", "title": "Reduced Matrix Multiplication: Input-Adaptive Matrix-Product Reduction for LLM Inference", "published": "2026-08-13T16:16:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "efficient-inference", "long-context", "multimodal-llm", "vision-language" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.13425", "title": "Motor, Cognitive, or Corpus? What Survives Cross-Lingual Transfer in Speech-Based Parkinsons Disease Detection", "published": "2026-08-13T16:15:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13420", "title": "Enhancing Virtual Agents through SLMs and Edge-Computing: An Exploratory Evaluation of Think and Memory Processes", "published": "2026-08-13T16:12:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13387", "title": "CROP: Task Relevance via Counterfactuals for Selective On-Policy Distillation", "published": "2026-08-13T15:48:16Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13343", "title": "AmalthAI: An Open-Source Computer Vision Platform for Cultural Heritage", "published": "2026-08-13T15:14:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13334", "title": "RippleMem: From Isolated Retrieval to Associative Recollection for Long-Term Agent Memory", "published": "2026-08-13T15:05:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13317", "title": "StateBridge: Training-free Hidden-state Alignment for Latent Communication in LLM Multi-Agent Systems", "published": "2026-08-13T14:40:59Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "llm-agent", "multi-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.13316", "title": "Foundation models for movement data: Are they ready for prime-time?", "published": "2026-08-13T14:39:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13309", "title": "How Good are Foundation Models in Longitudinal MRI Disease Progression Reasoning?", "published": "2026-08-13T14:35:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13292", "title": "Refine After Generation: Toward Correct and Concise Patches in LLM-based Program Repair", "published": "2026-08-13T14:25:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13267", "title": "How Do VLMs Behave When Blind or Misled? Behavioral Evaluation of VLMs on Scientific Figures", "published": "2026-08-13T14:06:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13262", "title": "Into the ORBIT for Time Series: Training Regimes for Foundation Models", "published": "2026-08-13T14:00:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13258", "title": "Self-Referential Induction Increases Response Instability Relative to Unresolvable and Verifiable Questions in Large Language Models", "published": "2026-08-13T13:58:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13239", "title": "Reasoning for Social Audio-Visual Question Answering: Where Do We Stand?", "published": "2026-08-13T13:44:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13226", "title": "CoverPrune: Coverage-Driven Token Pruning for 3D VLMs via Optimal Transport", "published": "2026-08-13T13:29:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13179", "title": "Teach the Magnitude, Not the Direction: Verifier-Bounded Credit Assignment for Multi-Turn Multi-step LLM Agents", "published": "2026-08-13T12:44:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13173", "title": "SkillShapley: Boundary-Adaptive Shapley Valuation for Skill Step Attribution in LLM Agents", "published": "2026-08-13T12:41:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13167", "title": "TRAPSBench: Vision-Language Models Encode but Fail to Express Epistemic Restraint", "published": "2026-08-13T12:32:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13120", "title": "SkillEvo: Self-Renewing Evolution Gradients from Multi-Turn Interaction Feedback", "published": "2026-08-13T11:49:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13119", "title": "QuISE: Defense against Typographic Attacks on VLMs via Query-Irrelevant Semantic Editing", "published": "2026-08-13T11:47:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13113", "title": "EgoMonth: A Month-Level Egocentric Video Benchmark for Long-Term Spatiotemporal Memory", "published": "2026-08-13T11:40:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13101", "title": "CASA: Content-Acoustic Speaking Assessment with Speech Encoder and Large Language Model", "published": "2026-08-13T11:25:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13031", "title": "UniTraffic-Agent: Unified Traffic Video Reasoning for AI City Challenge 2026 Track 3 with Two Out-of-Domain Evaluations", "published": "2026-08-13T10:00:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13029", "title": "Static analysis-guided agentic AI translation enables Rust as a full stack bioinformatics language", "published": "2026-08-13T09:58:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12995", "title": "OGR-MARL: Option-Guided Residual Multi-Agent Reinforcement Learning for Heterogeneous USV Cooperative Pursuit in Constrained Port Waterways", "published": "2026-08-13T09:19:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12990", "title": "LycheeMemory V2: Efficient Long-Term Memory for LLM Agents via Semantic Segment-Level Consolidation", "published": "2026-08-13T09:13:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12986", "title": "STAR: Structured Tokenization and Target-Aware Interest Representation for PCVR Prediction", "published": "2026-08-13T09:09:54Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "industrial-ranking", "priority-org-tencent", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.12984", "title": "Reconcile Once, Write Anytime: A Trust-Tiered Librarian and a Multi-Agent Writer for Drift-Free, Point-in-Time Research", "published": "2026-08-13T09:09:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12977", "title": "Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents", "published": "2026-08-13T08:57:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12975", "title": "DTAMLP: Denoise Time-aware MLP for Session-based Recommendation", "published": "2026-08-13T08:55:40Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12953", "title": "Unifying Depth and Width Pruning for LLMs via Binary Knapsack Optimization", "published": "2026-08-13T08:32:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12944", "title": "CardioState-JEPA: Delay-Aware Cross-Modal Learning of a Shared Cardiac Representation", "published": "2026-08-13T08:21:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12925", "title": "Momentum as Residual-Driven Multiplier Correction for Deep Learning Optimization", "published": "2026-08-13T08:04:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12921", "title": "Discovering Efficient and Explainable Communication Topologies for LLM-based Multi-Agent Systems via Causal Inference", "published": "2026-08-13T08:03:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12920", "title": "TennisVAR: A Stroke-Evidence-Grounded Multimodal Large Language Model for Tactical Reasoning in Tennis Videos", "published": "2026-08-13T08:01:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12911", "title": "Beyond Visual Evidence: Revealing and Mitigating Relational Privacy Leakage in Document MLLMs", "published": "2026-08-13T07:53:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12904", "title": "HounsWorld: A Multimodal World Model for Hidden Patient-State Readout, Reconstruction, and Simulation", "published": "2026-08-13T07:41:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12898", "title": "NaviDC-OCR: Navigating Document Parsing Across Digital and Camera-Captured Documents", "published": "2026-08-13T07:34:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12895", "title": "Agent Behavioral Contracts II: Certifying Compositional Reliability Without Assuming Independence", "published": "2026-08-13T07:25:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12888", "title": "When Your Agent Opens the Chat App: Agent-Controlled Search over Raw Chat Logs Rivals Structured Memory", "published": "2026-08-13T07:10:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14733", "title": "A Novel Fourier Feature Network for Solving Partial Differential Equations", "published": "2026-08-13T06:29:30Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12854", "title": "BrainWAM: Action-Space Coordination of Semantic Priors and Predictive Dynamics for Autonomous Driving", "published": "2026-08-13T05:56:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12852", "title": "Falsehood and Impossibility Are Different Directions in an AI's Representation of Language", "published": "2026-08-13T05:52:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12851", "title": "Practice Makes Unsafe: Skill Misevolution in Self-Improving LLM Agents", "published": "2026-08-13T05:47:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13626", "title": "A Calibrated Test of Internal Action Maps: State Signals Without Global Affine Closure", "published": "2026-08-13T05:41:24Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12847", "title": "Beyond Retrieval: Query-Conditioned Reuse of Long-Horizon Agent Trajectories", "published": "2026-08-13T05:39:49Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12845", "title": "FSGR: Mitigating Token Frequency Bias for Fair SID-Based Generative Recommendation", "published": "2026-08-13T05:34:51Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12843", "title": "Heterogeneous Vision-Language Ensemble with Disagreement-Aware Reranking for Text-Based Person Anomaly Retrieval", "published": "2026-08-13T05:28:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12841", "title": "AQuA: Recursively Self-Improving Quantitative Trading Research Agents", "published": "2026-08-13T05:25:42Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "agent-memory", "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12831", "title": "Fast A/B/n Testing: Exact Multi-Policy Comparison via Tree-Coupled Feedback Sharing", "published": "2026-08-13T05:10:43Z", "tracks": [ "post-training", "recommendation" ], "review_buckets": [ "industrial-fulltext-review", "query-collision" ], "matched_queries": [ "production-evidence", "recsys-general", "reward-model", "search-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B07", "plan_reason": "LLM 架构、长上下文、KV cache 与评测基础设施 — completed" }, { "arxiv_id": "2608.12821", "title": "HiRoute: Hierarchical Routed Prompt Tuning for Safety Alignment of Large Language Models", "published": "2026-08-13T04:49:51Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12812", "title": "A Comprehensive Empirical Evaluation of Vector Database Systems for Approximate Nearest Neighbor Search: Performance, Quality, and Resource Trade-offs", "published": "2026-08-13T04:41:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13625", "title": "Reward Machines for Signal Temporal Logic", "published": "2026-08-13T04:25:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12781", "title": "Beyond Correctness: Benchmarking and Aligning Response Behaviors in Hybrid-Thinking MLLMs", "published": "2026-08-13T03:44:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "vision-language" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12778", "title": "DrEM: Dual-Side Robust Ensemble Ranking from Noisy User Preference Predictions in Video Recommendation", "published": "2026-08-13T03:36:09Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "production-evidence", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.12771", "title": "Memorization Diagnostics for Code LLMs Should be Scale-Aware", "published": "2026-08-13T03:29:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12764", "title": "Beyond Outcome Rewards: Step-Level Self-Distilled Policy Optimization for Deep Search Agents", "published": "2026-08-13T03:13:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12753", "title": "Decentralized Multi-Player Q-Learning in Episodic Markov Decision Processes with Information Asymmetry", "published": "2026-08-13T03:06:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12746", "title": "Dual-Stream Cross-Anchor Correction Grounding Long-Form Captions and the Domain Limits of Object-Level Anchors", "published": "2026-08-13T02:52:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12745", "title": "A Cloud-Edge System for Multimodal Clinical Screening in Resource-Constrained Rural Settings", "published": "2026-08-13T02:45:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12743", "title": "Spatial Memory Agent: Experience-Grounded Procedure Memory for Spatial Intelligence", "published": "2026-08-13T02:42:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12720", "title": "ERSkill: Evolving for Skill-Guided Adaptive Memory Retrieval", "published": "2026-08-13T02:06:01Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13622", "title": "ARC: Fair Relative Advantage Comparison in Open-Ended Real-World Interaction", "published": "2026-08-13T01:51:53Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "reward-model", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12700", "title": "A Contract-Grade Verifier for LLM-Generated GPU Kernels, and a Native Blackwell Backward for the Gated-Linear-Recurrence Family", "published": "2026-08-13T01:25:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12677", "title": "The Role of Natural Language Understanding in Multimodal Video-Based Dengue Diagnosis", "published": "2026-08-13T00:27:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12630", "title": "Novels generated by language models show compressed formal variation", "published": "2026-08-12T22:32:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12626", "title": "LLMs Are Not Good Strategists, Yet Memory-Enhanced Agency Boosts Reasoning", "published": "2026-08-12T22:17:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12590", "title": "Auditable agentic AI for evidence-grounded thyroid ultrasound diagnosis and reporting", "published": "2026-08-12T21:02:59Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14728", "title": "Tail-Aware Top-$k$ On-Policy Distillation", "published": "2026-08-12T20:31:49Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13613", "title": "VoiceDesigner: Text-to-Voice Generation and Editing via Unified Diffusion Modeling and Data Augmentation", "published": "2026-08-12T18:04:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12476", "title": "Governed Persistent Memory: Source-Bound State Semantics and Fail-Closed Release for Long-Horizon Agents", "published": "2026-08-12T18:00:42Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12311", "title": "The Role Specialization Model (RSM): Coordinating LLM-Based Tools in Agentic Software Development - An Exploratory Case Study", "published": "2026-08-12T17:57:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12308", "title": "DreamFly: Causal Memory and Receding-Horizon Diffusion Planning for Aerial Vision-Language Navigation", "published": "2026-08-12T17:54:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12307", "title": "AI4AI at Test-Time: Strong-to-Weak Capability Transfer via Harnesses", "published": "2026-08-12T17:53:18Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12290", "title": "Beyond Trial-and-Error: Agentic Optimization for Image-to-Video Adherence", "published": "2026-08-12T17:35:16Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12282", "title": "VAKRA: Evaluating Multi-Hop Reasoning Across APIs and Retrieval Under Tool-Use Policies", "published": "2026-08-12T17:27:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12278", "title": "Structural Silence: When AI Infrastructure Fails Speakers of Underrepresented Languages", "published": "2026-08-12T17:17:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12269", "title": "A Cascaded Unsupervised-Supervised NLP Pipeline for Detecting Accusatory Language in Public Procurement", "published": "2026-08-12T17:09:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12262", "title": "Diagram-MMU: A Multi-Modal Benchmark for Scientific Diagrams", "published": "2026-08-12T17:04:13Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12253", "title": "One Frozen Simulator Is Not Enough: Simulator Collapse in Multi-Agent RL", "published": "2026-08-12T16:55:50Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12246", "title": "VICBench: A Multi-Language Benchmark for Code Vulnerability Detection", "published": "2026-08-12T16:45:49Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12219", "title": "ScreenShot: A Foundation Model for Few-Shot Combination Drug Screening", "published": "2026-08-12T16:13:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12218", "title": "Information Abundance Paradox: Long-Context Training Undermines Parametric Knowledge", "published": "2026-08-12T16:13:05Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12194", "title": "HYDRA: Hyperbolic Dynamic Representation Architecture for Kolmogorov-Arnold Networks", "published": "2026-08-12T15:48:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12192", "title": "How to Spend Your Oracle Budget: Practical Guidance for Protein Structure Prediction Models", "published": "2026-08-12T15:46:57Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12440", "title": "Specification-first convergence with an AI coding agent: a case study of dismantling a core architectural invariant across 189 files in a 717k-line codebase with no test oracle and no human code review", "published": "2026-08-12T15:35:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13612", "title": "SemPlan: Benchmarking Structured Semantic Planning for LLM-Based Queries over Enterprise Data", "published": "2026-08-12T15:31:54Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12138", "title": "A corpus-specific clinical RAG system matches or outperforms newer frontier LLMs on HealthBench", "published": "2026-08-12T14:55:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12121", "title": "QV-PIC: Query-Aware Visual Position-Independent Caching for Efficient RAG Serving", "published": "2026-08-12T14:40:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12099", "title": "RT-SEMamba: Real-Time Speech Enhancement Mamba via Progressive Knowledge Distillation", "published": "2026-08-12T14:21:05Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "model-compression", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.12436", "title": "Multi-AUV Ad-hoc network-based Target Tracking: A Value Gradient Guidance Multi-Agent Diffusion Reinforcement Learning Approach", "published": "2026-08-12T14:15:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12063", "title": "Learning Loco-Manipulation From SMPC Demonstrations With Sparse Offline-to-Online RL", "published": "2026-08-12T13:48:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14724", "title": "Privacy-Preserving Dataset Curation for Kuala Lumpur Urban Traffic: Grounded Vision-Language Detection with Spatial Vehicle-Context Filtering", "published": "2026-08-12T13:45:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12435", "title": "MARCH: Scaling Recurrent Memory with Content-Routed State Anchors", "published": "2026-08-12T13:45:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12036", "title": "Mechanist: AI as a Scientific Instrument for Discovering the Mechanisms of Intelligence", "published": "2026-08-12T13:19:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12018", "title": "Poly-Dialectal Neural Machine Translation System for Bangla Regional Dialects", "published": "2026-08-12T12:55:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12008", "title": "Asymptotic Risk Calibration for Selective Question Answering", "published": "2026-08-12T12:45:45Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11994", "title": "Claim-Level Reliability Assessment for Efficient Test-Time Reasoning", "published": "2026-08-12T12:33:05Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11981", "title": "Benchmarking Trustworthiness of SLMs: Pre-trained vs. Compressed", "published": "2026-08-12T12:14:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11980", "title": "Learning from Unreachable Rewards: Hint-Conditioned Reinforcement Learning for Generative Recommendation", "published": "2026-08-12T12:13:08Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11977", "title": "Retry, Switch, or Abstain? Learning Strategy-Aware Tool-Use Policies via Controlled Error Injection", "published": "2026-08-12T12:08:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12429", "title": "SynWeaver: Website-Prior Task and Trajectory Co-Synthesis for Web Agents", "published": "2026-08-12T12:03:20Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11973", "title": "Sci-Surf: Navigating Scientific Literature Discovery through Human Feedback and Intelligent Summarization", "published": "2026-08-12T12:02:17Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.11967", "title": "LoongReflect: Boosting Long-Horizon Reflection in Search Agents via Global Perspective Distillation", "published": "2026-08-12T11:56:03Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning", "agentic-rl", "llm-rl", "long-context", "multi-agent", "on-policy-distillation", "tool-agent" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B10", "plan_reason": "Agentic RL 与长时序 credit assignment — completed" }, { "arxiv_id": "2608.11965", "title": "Developing LLM-based Multi-Agent Systems in Software Engineering: A Mixed-Method Experience Report", "published": "2026-08-12T11:53:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11941", "title": "OEIS Open: How many conjectures can language models turn into theorems?", "published": "2026-08-12T11:28:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11924", "title": "Spark-to-Paper: End-to-End Research Paper Generation as a Composable Skill", "published": "2026-08-12T11:11:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11919", "title": "LazyTrain: Limited-resource Allocation toward Zero-waste Yield Optimization in Large Language Model Training", "published": "2026-08-12T11:04:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11909", "title": "Disentangling the Expressivity of RoPE", "published": "2026-08-12T10:37:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11907", "title": "A Model-Internal Protocol for Assessing Multimodal Models as Integrated Systems", "published": "2026-08-12T10:35:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11891", "title": "Benchmark-Based Comparative Assessment of Publicly Benchmarked Indian Foundation Models: A Capability and Evaluation-Maturity Framework", "published": "2026-08-12T10:19:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11859", "title": "Small-Scale Experiments: Are We There Yet?", "published": "2026-08-12T09:47:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11838", "title": "GeoBridge: Decoupled Semantic Conditioning for Generative Image Geolocalization", "published": "2026-08-12T09:21:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11829", "title": "Towards Understanding On-Policy Distillation through the Lens of Test-Time Scaling", "published": "2026-08-12T09:13:35Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11816", "title": "How China-Origin Vision-Language Models Move from Refusal to Reframing in State Alignment", "published": "2026-08-12T08:58:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11810", "title": "Can Vision Models Read the Radar Display? On the Feasibility of Radar Imagery for Air Traffic Complexity Estimation", "published": "2026-08-12T08:53:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11787", "title": "GRPO for Financial Advice Generation: Outperforming Commercial LLMs under CATE Evaluation", "published": "2026-08-12T08:28:15Z", "tracks": [ "post-training", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11786", "title": "Language-Conditional Dequantization: Recovering What Quantization Steals from Non-English Languages", "published": "2026-08-12T08:28:03Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11775", "title": "The Sleeping Agent: What Gist-Based Context Compression Loses and Why", "published": "2026-08-12T08:19:15Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11767", "title": "Causal Structure is Inducible but Functionally Decoupled: The Routing/Readout Boundary of a Typed Mechanism Library", "published": "2026-08-12T08:10:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11758", "title": "AWARe: Mitigating Catastrophic Forgetting via Activation-Weighted Adaptive REtention", "published": "2026-08-12T07:55:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11755", "title": "MuseCritic: Learning Multi-Aspect Song Rewards through Natural-Language Aesthetic Critiques", "published": "2026-08-12T07:49:17Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11753", "title": "LabelFusion-TS: Fusing Large Language Models, Transformer Encoders, and Financial Time Series for Monetary-Policy Stance Classification", "published": "2026-08-12T07:46:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12419", "title": "LoKiFormer: Locality-aware Attention with Decoupled Knowledge Memory for Efficient Large Language Model Pretraining", "published": "2026-08-12T07:45:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11749", "title": "MOON: Multi-Objective OrthoNormalized Updates for Multitask Learning", "published": "2026-08-12T07:41:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14721", "title": "AeroGround: A Comprehensive Benchmark for Aerial-Ground Collaborative Reasoning", "published": "2026-08-12T07:31:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11741", "title": "JieZi: A Large-Scale Expert-Audited Dataset and Benchmark for Ancient Chinese Character Exegesis", "published": "2026-08-12T07:30:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11738", "title": "Advancing MLLM-based UAV Image Understanding and Reasoning: A Benchmark and a Training-Free Multi-Agent System", "published": "2026-08-12T07:25:32Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11735", "title": "Locating and Controlling Implicit Personalization in Large Language Models", "published": "2026-08-12T07:18:31Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11715", "title": "When the API Speaks the Wrong Language: Revisiting Post-Training for Multilingual Tool Use", "published": "2026-08-12T06:55:04Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.11691", "title": "LEMUR: Latent Entropy-aware Multimodal Unlearning via Visual-anchored Reasoning Redirection", "published": "2026-08-12T06:03:50Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "multimodal-llm", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11683", "title": "FrontierFinance: A Challenging Benchmark for Measuring Frontier Intelligence of Finance Agents", "published": "2026-08-12T05:43:26Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11681", "title": "Learning from Multimodal Pseudo-Labels for Robust Open-Vocabulary Instance and Panoptic Segmentation", "published": "2026-08-12T05:39:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11669", "title": "Rubric Dropout: A Simple Way to Mitigate Reward Hacking in Rubric-as-Reward RL", "published": "2026-08-12T05:29:21Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B09", "plan_reason": "Rubric、外部 rollout 与多奖励 RL — completed" }, { "arxiv_id": "2608.11661", "title": "Low-Interaction-Rank Learning: Unifying Multiplicative Dual-Encoder Heads", "published": "2026-08-12T05:09:44Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11660", "title": "Hybrid-Policy Self-Editing for Composable Unstructured Knowledge Editing", "published": "2026-08-12T05:06:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11658", "title": "Is Per-Agent Policy Composition Safe? Rethinking Successor-Feature Transfer in Cooperative Multi-Agent Reinforcement Learning", "published": "2026-08-12T04:56:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11655", "title": "Motion-as-Prompt: Enhancing Motion Reasoning in Multimodal Large Language Models via Motion-Guided Cross-Frame Visual Prompting", "published": "2026-08-12T04:54:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11650", "title": "Confucius4-TTS: Transcript-Free Cross-Lingual Zero-Shot TTS with a Learnable Speaker Encoder", "published": "2026-08-12T04:48:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11643", "title": "Robustness of AI-Art Detectors under Generator Shift", "published": "2026-08-12T04:41:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11624", "title": "Learning to Persuade Exposes How Easily LLMs Abandon Correct Beliefs", "published": "2026-08-12T04:11:22Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11623", "title": "FM-LLM: A frequency-enhanced mixture-of-experts framework for adapting LLMs to time series forecasting", "published": "2026-08-12T04:09:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11616", "title": "MBA: Multimodal Benchmark and Agents for Real-World Business Ideation", "published": "2026-08-12T03:52:57Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14718", "title": "VideoGAIA: A Benchmark for General AI Assistants on Agentic Video Understanding", "published": "2026-08-12T03:24:44Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11604", "title": "Learning from Online User Feedback for Shopping Agents", "published": "2026-08-12T03:24:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11587", "title": "Robust Multi-Tier Infant-Centered Audio Understanding with Whisper via Structured Speaker Conditioning", "published": "2026-08-12T02:55:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11573", "title": "Reinforcing Step-level Reasoning for Effective Self-Correction in LLMs", "published": "2026-08-12T02:31:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11552", "title": "Beyond Single-Turn Confidence: Trajectory-Adapted Uncertainty Quantification for LLM Agents", "published": "2026-08-12T01:39:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11511", "title": "Let it Cook: Learning to Wait in Sequential Decision Making", "published": "2026-08-11T23:55:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11493", "title": "From Prompting to Behavioral Alignment: Personalized LLM Judges for Recommendation Evaluation", "published": "2026-08-11T23:06:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11469", "title": "The Next Challenge for Agentic Cybersecurity: A Realistic, Contamination-Free Reverse Engineering Benchmark", "published": "2026-08-11T22:14:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11458", "title": "Multi-Agent Target-Existence Verification and Learned Mask Geometry Refinement: Winning Report of the MeViS-Text Track at the 8th LSVOS Challenge 2026", "published": "2026-08-11T21:47:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11434", "title": "Benchmarking LLM Judges for Mobile Agent Evaluation", "published": "2026-08-11T21:00:46Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11426", "title": "Is Convergence Inevitable? Tracing Output Homogeneity Back to Base Models", "published": "2026-08-11T20:47:06Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "post-training", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11410", "title": "Unmasking Toxic Mimicry in Medical Offline Reinforcement Learning for ICU Sepsis Management via Counterfactual Clinical Audits", "published": "2026-08-11T20:17:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11408", "title": "Measure, Don't Optimize: Forecasting Recovery in LLM Unlearning", "published": "2026-08-11T20:16:20Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11403", "title": "When Self-Consistency Backfires: Majority Vote Hurts the Majority of Hard Science Problems for Small LLMs", "published": "2026-08-11T20:08:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11386", "title": "The Devil Is in the Interface: Evaluating How Tool Architecture Shapes Coding Agent Behavior", "published": "2026-08-11T19:50:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11381", "title": "From Numbers to Judgment: Specialist LLM Agents and Reinforcement Learning for European Listed Real Estate", "published": "2026-08-11T19:42:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11361", "title": "Lifecycle-Optimal Tokenization: Vocabulary Size as a Deployment-Regime-Dependent Infrastructure Parameter", "published": "2026-08-11T19:12:31Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11352", "title": "ODE-Based Transformer Decoders for Iterative Sign Language Translation", "published": "2026-08-11T18:58:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11350", "title": "Self-Evolving Embodied Agents via Skill-Harness Evolution", "published": "2026-08-11T18:55:58Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "reward-model", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.11349", "title": "Dynamics Models for Offline Hyperparameter Selection in Real-World RL", "published": "2026-08-11T18:54:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11343", "title": "Can Frontier LLMs Match Natively Multimodal Embeddings? A Comparison on Hard-Negative Text-to-Image Retrieval", "published": "2026-08-11T18:49:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11329", "title": "Qwen-MusicAVQA-7B: A Multimodal Model for Music Audio-Visual QA", "published": "2026-08-11T18:28:30Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11191", "title": "Test-Time Self-Evolving GUI Visual Grounding via Reflection-Guided On-Policy Self-Distillation", "published": "2026-08-11T17:50:25Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "on-policy-distillation", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.11167", "title": "MultiModal Code-Switching: Interleaving Visual Objects into Language for Explicit Object-Level Alignment", "published": "2026-08-11T17:28:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11166", "title": "Agentic Configuration Management (ACM): A Reference Configuration Model for Governed Agentic Systems", "published": "2026-08-11T17:28:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11292", "title": "Self-Evolving Code-with-Image Reasoning", "published": "2026-08-11T17:14:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11152", "title": "Scheduling Mixed RL Rollouts Beyond Prefix Locality", "published": "2026-08-11T17:10:50Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "efficient-inference", "llm-rl", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.11149", "title": "PRMU: A Corpus-Free Benchmark for Person-Centric Knowledge Unlearning in Multimodal Large Language Models", "published": "2026-08-11T17:09:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11143", "title": "A Recommendation System Approach for Interference-Robust Sensor Subset Selection", "published": "2026-08-11T17:01:44Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14712", "title": "Which Question Is Your Attention Metric Answering? Attention Rows as Compositional Data", "published": "2026-08-11T16:44:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11110", "title": "Actions Speak Louder than Words: Measuring Cross-Lingual Policy Retention in Tool-Using Agents", "published": "2026-08-11T16:18:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11095", "title": "Why Does CLAUDE.md Keep Growing? Catastrophic Remembering in Agentic Coding", "published": "2026-08-11T16:00:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11061", "title": "Batch Size or Negatives? A Selection Rule for Memory-Constrained Recommender Training", "published": "2026-08-11T15:29:01Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11049", "title": "Multiclass Sentiment Analysis for Identifying Political Viewpoints", "published": "2026-08-11T15:21:23Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11045", "title": "ReRound: Reconstructive Rounding to Resolve Midpoint Ambiguity in Calibration-Free LLM Quantization", "published": "2026-08-11T15:18:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11027", "title": "Mapping and Measuring the Behavioral Evolution of Large Language Models", "published": "2026-08-11T15:07:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10996", "title": "ConRub-Med: Reinforcement Learning with Consensus Rubrics for Open-Ended Medical Question Answering", "published": "2026-08-11T14:48:15Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10983", "title": "TimeRoute: Time-Aware Modality Routing and Diffusion for Multi-Modal Recommendation", "published": "2026-08-11T14:36:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon", "priority-org-tiktok" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10964", "title": "CARE: Confidence-Aware Reasoning for Reliable Medical VQA", "published": "2026-08-11T14:28:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10954", "title": "Evidence-Grounded Trustworthy Multimodal Reasoning and Evaluation Benchmark in Complex Urban Scenes", "published": "2026-08-11T14:23:07Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10949", "title": "StreamFlow: Dynamic Memory Flows for Streaming Video Understanding", "published": "2026-08-11T14:19:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13606", "title": "MobileMem: Learning from a Year of Mobile Experiences", "published": "2026-08-11T14:09:51Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10934", "title": "Understanding the Architecture of Coding Agents: An Exploratory Study Using a Research Prototype", "published": "2026-08-11T14:02:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10932", "title": "Temporally Grounded Compositional Camera Motion Understanding via Geometric Knowledge Distillation", "published": "2026-08-11T14:00:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10920", "title": "IO Factory: Simulating AI-Enabled Influence Campaigns at Scale", "published": "2026-08-11T13:43:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10908", "title": "Order Matters: LVLMs as Judges for Temporal Reasoning in Image Sequences", "published": "2026-08-11T13:29:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "multimodal-llm", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.10906", "title": "GitSkills: A Dataset of Agent Skills on GitHub", "published": "2026-08-11T13:28:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10897", "title": "Partially Observable Learning for Multi-Platform Dispatch Optimization", "published": "2026-08-11T13:20:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10875", "title": "VibeLifeBench: Can Your Life Agent Be Proactive and Persistent in a Living World?", "published": "2026-08-11T12:52:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10867", "title": "Can Bayesian Optimization Efficiently Find a Strong Single Expert in Neural Thickets?", "published": "2026-08-11T12:41:49Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10837", "title": "TACTICL: Task-Aware Compression of Tabular ICL Models", "published": "2026-08-11T12:03:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10827", "title": "MIRA: Medical Image Reflection for Agentic Diagnosis", "published": "2026-08-11T11:57:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10823", "title": "MoE Proxy Models for Low-Cost Failure Reproduction and Diagnosis in LLM RL Post-Training", "published": "2026-08-11T11:50:48Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10812", "title": "Reference-Free Post-Training of Open Large Language Models for Multilingual Machine Translation", "published": "2026-08-11T11:30:38Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.10806", "title": "Assessing Reliability of BERT-Based Models on Question Answering Tasks", "published": "2026-08-11T11:24:07Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10796", "title": "E$^3$mo-Bench: A Scalable Benchmark for Multimodal Evoked and Expressed Emotion Understanding via Bayesian Pairwise Alignment", "published": "2026-08-11T11:05:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10775", "title": "SkillLens: Visual Skill Cards for Retrieval-Augmented GUI Action Prediction and On-Policy Distillation", "published": "2026-08-11T10:28:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10758", "title": "Where To Look? : Causal Tracing of Vision Encoders in VLM", "published": "2026-08-11T10:17:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10756", "title": "Embodied Multimodal Grounding for Open-Vocabulary Mobile Manipulation via Semantic 3D Gaussian Splatting", "published": "2026-08-11T10:16:30Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10743", "title": "Mitigating Context Interference for Reliable and Efficient Search Agents", "published": "2026-08-11T10:00:42Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10720", "title": "Ex-Omni-2D: Expressive Omni-Modal Dialogue Models with Native Visual Presence", "published": "2026-08-11T09:37:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10714", "title": "Conversational Orchestration for Organic 6G", "published": "2026-08-11T09:32:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10706", "title": "MMArt: A Multi-Perspective Multimodal Dataset for Visual Art Understanding", "published": "2026-08-11T09:28:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10700", "title": "Deciding When to Rely on Visual Information: Gated Multimodal Fusion in Sequential Recommendation", "published": "2026-08-11T09:21:20Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10692", "title": "SPIEval: Evaluating Large Language Models as Mobile Assistants over Scattered Personal Information", "published": "2026-08-11T09:14:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10676", "title": "Self-Correcting Long-Horizon Search Agents via Tree-Structured Memory", "published": "2026-08-11T08:56:42Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10665", "title": "VERDICT: Training-Free Step-Wise Verification of Multimodal Reasoning via Disagreement-Aware Consensus", "published": "2026-08-11T08:46:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10636", "title": "DistilVDR: A Compact End-to-End Visual Document Retriever via Dual-Student Distillation", "published": "2026-08-11T08:23:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10628", "title": "InSight-doc: Agentic Visual Perception for Long-Document Understanding", "published": "2026-08-11T08:15:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10627", "title": "Decomposition-Induced Context-Memory Conflict: When Fact-Checking Pipelines Contradict Their Own Source Text", "published": "2026-08-11T08:15:05Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10626", "title": "Dual-Loop Self-Evolution via Verifiable Emotion Feedback for Multi-Turn Empathetic Dialogue", "published": "2026-08-11T08:14:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10622", "title": "A Study of Cursorrules Files in GitHub Open Source Projects", "published": "2026-08-11T08:08:49Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10621", "title": "ProbGuard: Calibrated Safety Risk Estimation from LLM Output Distributions", "published": "2026-08-11T08:08:41Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10605", "title": "Compute-Optimal Is Not Cluster-Optimal: Systems-Aware Scaling for Sparse Mixture-of-Experts", "published": "2026-08-11T07:49:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10538", "title": "SKILLER: Language-Level Reinforcement Learning for Reusable Skill Extraction in Small Language Models", "published": "2026-08-11T06:22:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10532", "title": "Benchmarking LLM-Guided Control-Plane Policies for Backend Fault Isolation in HAProxy", "published": "2026-08-11T06:15:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10530", "title": "On Understanding, Identifying, and Mitigating Vulnerabilities in Agentic Large Language Models", "published": "2026-08-11T06:11:26Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agent-planning", "multi-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.10522", "title": "Unlocking the Power of Medical Tabular Data via Semantic-Aware Multimodal Pre-training", "published": "2026-08-11T05:57:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10513", "title": "SafeCap: Improving LVLM Safety with Image Captioning Reinforcement Learning", "published": "2026-08-11T05:37:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10509", "title": "MAP-Graph: Provenance-Aware Shared Memory for Multi-Agent Workflows", "published": "2026-08-11T05:31:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10502", "title": "From Faulty Memories to Corrected Actions: Dependency-Guided Rollback Repair for Memory-Augmented Agents", "published": "2026-08-11T05:19:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10497", "title": "SapiensID 2.0: Aligning Human Recognition Foundation Models with Human Perception", "published": "2026-08-11T05:13:05Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10474", "title": "Stay or Stray - A Dynamical Systems Viewpoint of Popularity Bias", "published": "2026-08-11T04:41:24Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10462", "title": "Calibrating Post-Training Feature Shifts for LLM Data Contamination Detection", "published": "2026-08-11T04:18:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10450", "title": "Persistent Recursive Worlds Enable Autonomous Software Evolution", "published": "2026-08-11T04:04:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10447", "title": "Towards Efficient Reasoning in LLM-Based Recommender Systems via Model Merging", "published": "2026-08-11T04:01:35Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.10444", "title": "From Reasoning Depth to Reasoning Breadth: Evaluating Multi-Point Associative Reasoning in Large Language Models", "published": "2026-08-11T03:58:23Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14707", "title": "Semantic Uncertainty-Guided Orchestration in Hierarchical Multi-Agent Systems", "published": "2026-08-11T03:55:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10441", "title": "Detecting an Effect Is Not Learning to Act on It: A Reward-SNR Floor for LLM Acquisition Agents", "published": "2026-08-11T03:46:23Z", "tracks": [ "post-training", "recommendation" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-recommendation", "priority-org-amazon", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10430", "title": "Actionable Hallucination Detection: Translating Latent Uncertainty into Agentic Critique", "published": "2026-08-11T03:26:16Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-architecture", "tool-agent", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.10408", "title": "VisEditBench: Can Vision-Language Models Edit Visualization Code from Multimodal Feedback?", "published": "2026-08-11T02:52:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10402", "title": "TideRL: Boosting Agentic RL Goodput with Readiness-Aware Scheduling", "published": "2026-08-11T02:49:33Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10372", "title": "Invertible Logits Transformation for Accuracy-Preserving Post-Hoc Uncertainty Calibration", "published": "2026-08-11T02:08:05Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10366", "title": "DSAgentBench: Can Agents Automate End-to-End Data-Science Workflows in Real Computer Environments?", "published": "2026-08-11T01:45:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10363", "title": "Nutrition Data Infrastructure for the AI Era: Operationalizing FAIR for Agent-Mediated Research", "published": "2026-08-11T01:42:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10359", "title": "VoxSumm: A Multilingual Corpus of Long-Form Spoken News for Joint Summarization and Translation", "published": "2026-08-11T01:33:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10327", "title": "Toward a Theory of Value in AI Alignment", "published": "2026-08-10T23:57:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10319", "title": "Do Personalized Skills Help Coding Agents? An Empirical Study of Developer Interaction Histories", "published": "2026-08-10T23:41:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10297", "title": "Neural Tree Collaborative Filtering: Rethinking Graph Collaborative Filtering as Tree Collaborative Filtering with Curvature-Aware Propagation Depth", "published": "2026-08-10T23:03:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10296", "title": "Cracks in the Foundation: Seemingly Minor Architectural Choices Impact Long Context Extension", "published": "2026-08-10T23:03:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "long-context", "pretraining-data", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B07", "plan_reason": "LLM 架构、长上下文、KV cache 与评测基础设施 — completed" }, { "arxiv_id": "2608.10288", "title": "Power law graph attention: exact generalization of scaled dot-product attention, empirical collapse at inference", "published": "2026-08-10T22:45:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14705", "title": "On Cross-Validation for Hyperparameter Optimization of Deep Learning Image Classifiers", "published": "2026-08-10T22:44:42Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10273", "title": "Locally Deployable Small Language Models for Emergency Department Decision Support: A Systematic Benchmark of Fine-Tuning Strategies", "published": "2026-08-10T22:15:21Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10256", "title": "CRHT: A Continuous Regression Hybrid Transformer for Vessel Trajectory Prediction with Online Cluster Sampling", "published": "2026-08-10T21:42:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10240", "title": "Sequential Modality Dropout for Robust Multi-Modal Sequential Recommendation", "published": "2026-08-10T21:18:25Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10218", "title": "Mind Viruses: Self-Propagating Ideas in Multi-Agent LLM Systems", "published": "2026-08-10T20:37:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10196", "title": "ELMER: Evolutionary Language Model that Explores and Refines", "published": "2026-08-10T20:14:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10182", "title": "From Prediction to Incrementality: Causal Optimization for Large-Scale Targeting and Recommendation", "published": "2026-08-10T19:54:08Z", "tracks": [ "foundation-model", "post-training", "recommendation" ], "review_buckets": [ "industrial-fulltext-review", "p2-deferred-review", "query-collision" ], "matched_queries": [ "opd", "production-evidence", "recsys-general", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.10178", "title": "One Recipe, Many Harnesses: What Self-Evolution Encodes Across Languages and Models", "published": "2026-08-10T19:45:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10154", "title": "Multimodal Item Parameter Estimation using Simulated Response Probabilitie", "published": "2026-08-10T19:15:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10149", "title": "REATS: LLM Reasoning-based Ensemble Learning for Adaptive Time Series Forecasting", "published": "2026-08-10T19:04:46Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "process-reward", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10126", "title": "Procedural Fairness Failures in RLHF from Preference Averaging", "published": "2026-08-10T18:38:16Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "preference-optimization", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.10120", "title": "ChronoSSM: Training for Temporally Aware Representations in Autoregressive State Space Models", "published": "2026-08-10T18:35:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10090", "title": "CHORUS: Complementary Experts for High-Coverage Testbench Stimulus Generation", "published": "2026-08-10T18:02:42Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.09931", "title": "Perception Before Supervision: Self-Contained Visual Distillation from Counterfactual Blind Spots", "published": "2026-08-10T17:59:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09928", "title": "Multimodal Model Diffing for Feature Discovery and Control", "published": "2026-08-10T17:59:30Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "multimodal-llm", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09907", "title": "DistMoE: Private-data Rehearsal-free Routing in Mixture-of-Experts for Distributed Instruction Tuning", "published": "2026-08-10T17:52:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09900", "title": "Decoding-Level Taboo: A Diagnostic Stress Test for LLM Robustness", "published": "2026-08-10T17:47:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09885", "title": "SHE: Trajectory-driven Safety Harness Evolution for LLM Agents", "published": "2026-08-10T17:35:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09880", "title": "Financial Numerical Prediction and Allocation as Token Generation", "published": "2026-08-10T17:33:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09867", "title": "Stealing Reasoning Traces from Proprietary LLM APIs", "published": "2026-08-10T17:24:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09853", "title": "RynnValue: Scaling Robotic Value Foundation Models with Temporal Distance", "published": "2026-08-10T17:09:37Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09834", "title": "RA-FinBERT: Rule-aware LoRA adaptation for low-resource financial sentiment classification", "published": "2026-08-10T16:52:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09826", "title": "Distill Skills into Weights, Not Prompts: Abstract Skills as Privileged Signals for On-Policy Self-Distillation", "published": "2026-08-10T16:43:18Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09819", "title": "Macaron-V1: Towards Open Continual Learning with Self-Improvement and Mixture-of-LoRA", "published": "2026-08-10T16:39:55Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "long-context", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09802", "title": "SWE-Bench ProMax: Benchmarking Agents on Large-Scale Multilingual Code Refactoring", "published": "2026-08-10T16:23:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09799", "title": "SpecPath: Testing Coding Agents Across Contract-Equivalent Specification Histories", "published": "2026-08-10T16:19:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09789", "title": "ADOPD: Reference-Privileged On-Policy Distillation for MLLM-Based Industrial Anomaly Detection", "published": "2026-08-10T16:12:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09772", "title": "PragMatch: Separating Pragmatic Incongruity from Cross-Modal Mismatch in Large Vision-Language Models", "published": "2026-08-10T16:00:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09765", "title": "REFRAMED: Towards Realistic Audio Description Generation for Movies", "published": "2026-08-10T15:55:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09764", "title": "MoNo: Multiscale Optimal Transport Neural Operator for Solving PDEs on General Geometries", "published": "2026-08-10T15:55:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09745", "title": "SR-OPSD: Self-Referenced On-Policy Self-Distillation", "published": "2026-08-10T15:40:07Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation", "reward-model" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B08", "plan_reason": "OPD 与多教师/过程蒸馏 — completed" }, { "arxiv_id": "2608.09703", "title": "Matryoshka Language Model Suites", "published": "2026-08-10T15:07:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09696", "title": "Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models", "published": "2026-08-10T14:59:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09666", "title": "Open Evaluation Agent: Efficient and Promptable Evaluation of Visual Generative Models", "published": "2026-08-10T14:42:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09634", "title": "IntHQ: Task-Interactive Hierarchical Query on Dual-Stream Representations for Generative Recommendation", "published": "2026-08-10T14:13:47Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "generative-recommendation" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B01", "plan_reason": "8 月工业生成推荐与多模态 — completed" }, { "arxiv_id": "2608.09628", "title": "Satellite Trajectory Optimization via Proximal Policy Optimization for Space Debris Avoidance", "published": "2026-08-10T14:09:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09617", "title": "Bayesian Symbolic Regression with Entropic Reinforcement Learning", "published": "2026-08-10T13:56:45Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09605", "title": "TSPORec: Token Selection via Preference Optimization for LLM-Based Sequential Recommendation", "published": "2026-08-10T13:47:46Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.09568", "title": "Se-DPO: Self-Evolving Token Credit for Direct Preference Optimization", "published": "2026-08-10T13:05:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10045", "title": "Finding the Signal in the Spam: Jointly Learning Rewards and Worker Reliability from Pairwise Comparisons", "published": "2026-08-10T13:01:48Z", "tracks": [ "post-training", "recommendation" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "priority-org-amazon", "recsys-general", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09558", "title": "Training-Free Universal Approximation by Prompting Random Transformers", "published": "2026-08-10T12:57:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09555", "title": "Bidirectional Context Self-Distillation for Reinforcement Learning of Skill-Based LLM Agents", "published": "2026-08-10T12:53:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09548", "title": "ELBench: A Multi-Dimensional Benchmark for Education-Facing Large Language Models", "published": "2026-08-10T12:46:58Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09510", "title": "Build it, Break it, Repeat: Benchmarking and improving LLM-manipulated disinformation detection in social media posts", "published": "2026-08-10T12:13:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09507", "title": "Learning Preference Adaptation for Large Language Model Personalization via Verbal Reinforcement Learning", "published": "2026-08-10T12:11:47Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agentic-rl", "llm-rl", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09474", "title": "FaLCon: Facet-Anchored Retrieval with Late Consensus for Sim2Real Text-Based Person Anomaly Search", "published": "2026-08-10T11:42:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09467", "title": "RecoverFly: A Failure-Aware Reinforcement Learning Post-Training Framework for Aerial Vision-Language Navigation", "published": "2026-08-10T11:37:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09444", "title": "Depth-adaptive Inference of Looped Language Models via Continuous Depth Batching", "published": "2026-08-10T11:20:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09443", "title": "Coupled Graph--Policy Distillation for Personalized Medication Safety in Older Adults with Multimorbidity", "published": "2026-08-10T11:19:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09432", "title": "ZetaGPT: A Reference Implementation of Positional--Encoding--Free State--Space--Attention Language Models", "published": "2026-08-10T11:05:16Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "llm-rl", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09417", "title": "Why Post-Norm Transformers Collapse: Attention Amplification and Gradient Repair Failure", "published": "2026-08-10T10:45:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09380", "title": "OpenLoopEvolve: A Verifiable Self-Evolution Framework for Loop Policies in Long-Horizon Complex Tasks", "published": "2026-08-10T09:57:26Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agent-planning", "tool-agent" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B11", "plan_reason": "Agent 记忆、工具规划与自进化系统 — completed" }, { "arxiv_id": "2608.09374", "title": "CircuitReason-1k: Benchmarking Long-Horizon Visual-to-Symbolic Reasoning inElectrical Circuits", "published": "2026-08-10T09:55:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09351", "title": "Test-Time Augmentation for LLMs: When Input Diversity Beats Output Diversity at Matched Compute", "published": "2026-08-10T09:31:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09344", "title": "Beyond Global Editing: Per-Instance Disentangled Subspaces for Training-Free Hallucination Mitigation in LVLMs", "published": "2026-08-10T09:21:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09290", "title": "OpenCodeReview: Determinism over Non-Determinism for Cost-Effective Agent-Based Code Review", "published": "2026-08-10T08:43:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09278", "title": "Software Engineering for and with GUI Agent", "published": "2026-08-10T08:33:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09277", "title": "P$^{3}$: Joint Program-and-Proof Planning for Verified Code Generation", "published": "2026-08-10T08:33:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09276", "title": "Verifiably grounded machine interpretation of lunar geology", "published": "2026-08-10T08:32:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09273", "title": "Entropy-based Code Adversarial Translation for Real-world Repository Migration", "published": "2026-08-10T08:29:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09253", "title": "SkillSentry: Reliable Skill Execution for LLM Agents via Runtime Assurance", "published": "2026-08-10T08:13:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09251", "title": "MoRSE: Task-Oriented Multi-Agent System with Mixture of Role-Subtask Experts", "published": "2026-08-10T08:11:24Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10037", "title": "DOCSCHISEL: Adaptive Tool Documentation Optimization Framework for LLM Agents", "published": "2026-08-10T08:05:45Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09217", "title": "Beyond Solvability: Task Learnability as a Static Prior for LLM RL Post-Training", "published": "2026-08-10T07:43:05Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.09209", "title": "UNMASK: Discovering and Causally Verifying Spurious Shortcuts in Text Classifiers", "published": "2026-08-10T07:31:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09189", "title": "EmoS: A Theory-Grounded Framework for Evaluating and Aligning Emotional Intelligence in Spoken Language Models", "published": "2026-08-10T06:58:30Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09185", "title": "SiriusDeliver: Automating Data Warehouse Delivery at Tencent", "published": "2026-08-10T06:54:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09184", "title": "Agentic Router: An Execution-Grounded Continual Learning Approach With Memory", "published": "2026-08-10T06:53:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09145", "title": "Right Answer, Wrong Heat: Explanation-Aware Evaluation and Thermal-Grounded Feedback for MLLMs on Infrared Images", "published": "2026-08-10T05:49:33Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09142", "title": "An Agentic Generative Large Language Model for Treatment Planning of Colorectal Cancer", "published": "2026-08-10T05:42:03Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09130", "title": "MARA: Flow-Matching-Guided Multi-Agent Resource Allocation for Computational Resource Efficient Learning", "published": "2026-08-10T05:13:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09128", "title": "Social Gym and SPaRTan: Benchmarking and Improving LLM Social Reasoning via Multi-Agent Game Tournaments", "published": "2026-08-10T05:12:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09122", "title": "Visual Distortion Detection in UGC Images Using Large Multimodal Models", "published": "2026-08-10T05:00:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09119", "title": "Motif 3: Technical Report", "published": "2026-08-10T04:53:05Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agentic-rl", "software-agent", "tool-agent", "web-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.09097", "title": "SI-Edit: Toward Sketch-Instruction Guided Local Image Editing with Pixel-Level Precision", "published": "2026-08-10T03:50:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09072", "title": "A Unified Issue Resolution Benchmark for Requirement Clarification, Planning, and Code Generation for Coding Agents", "published": "2026-08-10T03:22:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09068", "title": "Pseudo2CodeQA: A Benchmark for LLM-Based Structured Algorithmic Reasoning in Code Generation", "published": "2026-08-10T03:17:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09053", "title": "Diagnosing as Cardiologists Do: ECG Agents with Doctor-Grounded Priors for Clinical Reasoning Across Diseases and Populations", "published": "2026-08-10T03:00:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09031", "title": "HOPPER: Learnable Hop Extraction for Linearized Graph Sequence Models", "published": "2026-08-10T02:31:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09016", "title": "PreGress: Ranking-Native Pre-training and Prompting for Graph Node Ranking", "published": "2026-08-10T02:16:37Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09011", "title": "Dynamic Distribution-Aware Uncertainty Tracking in Vision-Language Representation Learning", "published": "2026-08-10T01:53:28Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08975", "title": "How Can Rhetoric Reward-Hack AI Reviewers? Dissecting Rhetorical Sensitivity in AI-Based Peer Review", "published": "2026-08-10T00:42:51Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08968", "title": "GALA: Graph-Augmented LLM Agents for Root Cause Analysis and Incident Response in Microservices", "published": "2026-08-10T00:15:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08961", "title": "Gradient Under Microscope: Benchmarking Resource Utilization of Memory-Efficient Gradient Computation Methods", "published": "2026-08-09T23:41:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08960", "title": "Reading is not Reasoning: Bridging the Agentic Policy Gap in Vision-Text Compression", "published": "2026-08-09T23:38:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08950", "title": "Independent Patch Verification for Coding Agents with a Bidirectional Reconstruct-and-Verify Framework", "published": "2026-08-09T22:59:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08947", "title": "Can Webcam Gaze Constrain Mesa-Objectives in Driving Models? An Instrument Precision Analysis", "published": "2026-08-09T22:55:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08915", "title": "Investigating Multimodal Informativity under Different Partner Visibility Conditions in Video-Mediated Dialogue", "published": "2026-08-09T21:08:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08904", "title": "From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability", "published": "2026-08-09T20:31:52Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08878", "title": "DistillCache: KL-Guided Adaptive KV-Cache Eviction for Memory-Efficient LLM Inference", "published": "2026-08-09T19:33:43Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-rl", "long-context", "model-compression", "reward-model" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B07", "plan_reason": "LLM 架构、长上下文、KV cache 与评测基础设施 — completed" }, { "arxiv_id": "2608.08872", "title": "Approximation Rates for Metaplectic Neural Networks", "published": "2026-08-09T19:27:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08853", "title": "Beyond Routing: Decoupling Expert Dispatch and Aggregation in Sparse Mixture-of-Experts", "published": "2026-08-09T18:31:16Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-architecture", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08847", "title": "Explicit Boundary Markers for Subword Vocabularies", "published": "2026-08-09T18:16:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08809", "title": "Tevatron-Elastic: A Unified Abstraction for Training Elastic Retrievers and Rerankers", "published": "2026-08-09T16:45:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08804", "title": "ML-Based Hierarchical Prediction for Practical Energy Scheduling in Dynamic NTN-WPT Systems", "published": "2026-08-09T16:41:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08795", "title": "Toward Metacognitive One-Shot Indirect Prompt Injection: Strategy Abstraction Via Outcome-Conditioned Reflection", "published": "2026-08-09T16:19:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08793", "title": "Evidence-Calibrated Runtime Reconstruction for Agent Skills Across Heterogeneous Coding Agents", "published": "2026-08-09T16:18:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08775", "title": "OmnilingualGAIA2: Evaluating the Multilingual Gap in Frontier AI Agents", "published": "2026-08-09T15:49:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08764", "title": "Learning from Consensus and Disagreement: Unsupervised On-Policy Self-Distillation with Minority-Trajectory Contrast", "published": "2026-08-09T15:23:25Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08744", "title": "Can We Optimize the Performance-Carbon Emission Break-Even Point?: The Quest for Greener LLMs", "published": "2026-08-09T14:47:35Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "query-collision" ], "matched_queries": [ "efficient-inference", "llm-architecture", "model-compression", "post-training" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.08743", "title": "A Distribution Mapping Approach to Counterfactually Fair Reinforcement Learning", "published": "2026-08-09T14:42:47Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08675", "title": "Efficient Test-Time Scaling for LLM-based Time Series Forecasting", "published": "2026-08-09T12:45:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08654", "title": "The Scaffolding Matters More Than the Interface: A Controlled Comparison of MCP and CLI Tool Use Across Seven Agent Scaffoldings, Five Language Models, and One Software Task", "published": "2026-08-09T11:54:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08652", "title": "LegoLM: Structured Weight Sharing for Large Language Models", "published": "2026-08-09T11:51:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08650", "title": "The Evolution of Mixture-of-Experts Architectures in Large Language Models: Routing, Topology, Load Balancing, and Expert Parallelism", "published": "2026-08-09T11:46:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08634", "title": "Can Open-Weight Models Compete on Financial Text Comprehension?", "published": "2026-08-09T10:51:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08630", "title": "VLZip: Unified Visual and Textual Compression for Interleaved Long-Context Modeling", "published": "2026-08-09T10:31:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10021", "title": "Position Encoding in Transformers: From Absolute and Relative Methods to Rotary Position Embeddings and Long-Context Scaling", "published": "2026-08-09T10:15:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08605", "title": "ForestBench: A Unified Graph Framework for Evaluating Multi-Agent Collaboration", "published": "2026-08-09T09:38:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08604", "title": "Multi-Agent Reinforcement Learning via Agent-Specific Preference", "published": "2026-08-09T09:38:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08596", "title": "Goal-oriented Navigation Instruction Generation with Tour Video Priors", "published": "2026-08-09T09:21:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08583", "title": "Structure-Preserving Projection for Mitigating Modality Bias in LLM-Based Sequential Recommendation", "published": "2026-08-09T08:56:54Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08570", "title": "FailForge: Distilling Procedural Competence from Persistent Failures into Code Agents", "published": "2026-08-09T08:22:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08557", "title": "OpenVisTool: An Open Recipe for Synthesizing Instructive Visual Tool-Use Trajectories", "published": "2026-08-09T08:01:05Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08553", "title": "MotionCraft: Latent World Modeling with Sparse Attention for Visual Upscaling", "published": "2026-08-09T07:54:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08545", "title": "Curriculum Generation under Structured Parametric Environments for Robust Navigation Policies", "published": "2026-08-09T07:45:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08510", "title": "From Speech to Interaction: Analyzing Multimodal Systems in Cocktail-Party Scenarios", "published": "2026-08-09T06:11:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08506", "title": "Understanding Calibration and Truncation Error Propagation in Training-Free Low-Rank Compression for LLMs", "published": "2026-08-09T06:03:48Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "model-compression", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08477", "title": "VectraYX-Vision-1B: A Sub-2B Spanish/LATAM Cybersecurity Vision-Language Model with Structured Visual Reasoning and Native Tool Use", "published": "2026-08-09T04:46:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08443", "title": "Private Etymology: Designing Relational Reuse of Shared Symbols in Long-Term Human-AI Interaction", "published": "2026-08-09T03:23:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08422", "title": "Population-Level Generative Modeling for Ranking Data", "published": "2026-08-09T02:36:49Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08418", "title": "Learning Deep Modality-Shared Self-Expressiveness for Image Clustering with Textual Information", "published": "2026-08-09T02:23:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08417", "title": "Personalized Communication Skills for Agentic Recommender Systems", "published": "2026-08-09T02:23:45Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08413", "title": "Tangent: An Empirical Study of Testing Practices for LLM-Based Agent Applications", "published": "2026-08-09T02:08:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08392", "title": "CAP: A Scalable Benchmark for Evaluating Cross-Site Browser Agents with Complex Actions and Perception", "published": "2026-08-09T01:08:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08344", "title": "PRISM: A Predictive Protocol for Permutation Optimization via Landscape Diagnostics", "published": "2026-08-08T21:52:04Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-architecture", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08311", "title": "Ouroboros: A Self-Developing Frontier Coding Agent with Reviewed Core Evolution", "published": "2026-08-08T19:45:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08307", "title": "Frequency-Domain Dual-Branch Fusion for Medical Visual Question Answering", "published": "2026-08-08T19:36:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08300", "title": "Mitigating Over-Personalization in LLMs via Structured Memory", "published": "2026-08-08T19:23:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08282", "title": "Stateful CARS: Exact Cross-History Reuse for Policy-Constrained LLM Agents", "published": "2026-08-08T18:26:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08264", "title": "OBLIVION: Workflow-Level Operational Skill Unlearning for Deployed Agents", "published": "2026-08-08T17:51:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08255", "title": "Learning from Environmental Feedback: Credit Assignment across Multiple Timescales for Agentic Reinforcement Learning", "published": "2026-08-08T17:32:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10016", "title": "Sheaf-Based Federated Representation Learning", "published": "2026-08-08T16:38:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08220", "title": "Metanormative Theory for RL-Based Moral Agents", "published": "2026-08-08T16:33:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08219", "title": "VTO: Visual Tool Orchestration for Video Anomaly Detection", "published": "2026-08-08T16:28:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08188", "title": "Quantization Degradation in Large Language Models: A Signal-Noise Perspective", "published": "2026-08-08T15:28:45Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08176", "title": "Matching Supervision to the Student's Learning Capacity: A Unified Framework for On-Policy Self-Distillation", "published": "2026-08-08T15:13:07Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08167", "title": "Wiener Representation Filtering for VLM Hallucination Suppression", "published": "2026-08-08T14:51:17Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08164", "title": "STEMMA: An Adversarial Multi-Agent Framework for Evaluating Self-Identity Consistency in LLMs", "published": "2026-08-08T14:46:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08156", "title": "A Hybrid Nested Harness for Decoupling Structure and Parameters in LLM-Driven Optimization", "published": "2026-08-08T14:35:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08148", "title": "DoGMA: A Central-Dogma-Guided Foundation Model for Multi-Omics Alignment and Multi-Task Learning in Oncology", "published": "2026-08-08T14:15:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08131", "title": "Compositional Threat Analysis of Latent Compromise in LLM Agent Systems: The Order 66 Scenario", "published": "2026-08-08T13:35:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08127", "title": "Improving Constraint Models with LLM Agents", "published": "2026-08-08T13:22:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08107", "title": "NeuPAT: Neuron-aware Plasticity Allocation Tuning for Language-Preserving MLLMs", "published": "2026-08-08T12:40:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08100", "title": "Defending Retrieval-Augmented Intrusion Detection Against Knowledge Poisoning and Prompt Injection", "published": "2026-08-08T12:29:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08081", "title": "RotaryQuant: Fitting 120B MoE Models on Consumer Hardware via Fused Compressed-Space Attention", "published": "2026-08-08T11:58:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "model-compression", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.08067", "title": "DialectS2S: End-to-End Speech Dialogue Modeling for Low-Resource Chinese Dialects", "published": "2026-08-08T11:11:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08060", "title": "ZOMP: Zeroth-Order Multi-Modal Prompt Tuning for Vision-Language Models", "published": "2026-08-08T10:55:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08055", "title": "SodaMem: Evidence-Grounded Temporal Graph Memory for LLM Agents", "published": "2026-08-08T10:42:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08038", "title": "Stateful Multi-Agent LLMs for Cross-View Interface Alignment in Automotive Model-Based Systems Engineering", "published": "2026-08-08T09:49:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08032", "title": "Decided Upstream, Written Late: Locating and Pricing the Cross-Lingual Refusal Circuit of a Multilingual MoE", "published": "2026-08-08T09:39:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.08002", "title": "Evaluator Ensembles Under Reward Hacking: Covariance Geometry and Finite-Search Guarantees", "published": "2026-08-08T08:28:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07998", "title": "Give the Long-tail More SPACE: Promoting Provider Fairness in Next POI Recommendation", "published": "2026-08-08T08:20:21Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07989", "title": "PushDualGen: Enabling LLMs to Generate Semantic IDs with Interpretable Copy for Industrial Push Recommendation", "published": "2026-08-08T07:56:55Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-kuaishou", "production-evidence", "recsys-general" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B01", "plan_reason": "8 月工业生成推荐与多模态 — completed" }, { "arxiv_id": "2608.07987", "title": "Advantage-Guided Gate: Reshaping Open-Ended Reasoning for Vision-Based Spatial Intelligence", "published": "2026-08-08T07:46:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07978", "title": "Verication-driven closed-loop multi-agent large language modelframework for code-compliant structural design", "published": "2026-08-08T07:26:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07955", "title": "Self-Evolving Neuro-Symbolic Skills for Tool-Augmented Spatial Reasoning", "published": "2026-08-08T06:36:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07952", "title": "Persistent Semantic Entities in Tool-Augmented LLM Systems", "published": "2026-08-08T06:31:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07941", "title": "LAD-COD: Language-Aligned Dense Perception for Camouflaged Object Detection", "published": "2026-08-08T06:02:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07935", "title": "Adaptive Supervised Anchoring for On-Policy Self-Distillation", "published": "2026-08-08T05:46:32Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07933", "title": "EvoTrustRAG: Evolution-Aware Conflict Attribution and Evidence Handling for Reliable Retrieval-Augmented Generation", "published": "2026-08-08T05:40:03Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07932", "title": "SportsGrounder: Proposal-Aided Interleaved Grounding for Dense Sports Video Reasoning", "published": "2026-08-08T05:39:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07920", "title": "Forged Peer Judgments Mislead Multimodal LLM Judge Panels: Source-Blind Anchoring and Panel-Consensus Verification", "published": "2026-08-08T04:55:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07915", "title": "SPECTRA: Pushing the KV Cache Beyond the 2-Bit Cliff via Spectral Transform Coding", "published": "2026-08-08T04:43:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07899", "title": "TelemetrySuffBench: Is Agent Telemetry Sufficient for Failure-Origin Diagnosis?", "published": "2026-08-08T03:54:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07851", "title": "TEMPER: Tensorized Efficient Manifold-constrained Parameterization for Expressive Residual Routing", "published": "2026-08-08T01:45:27Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07822", "title": "Classical $\\mathrm{SU}(2)$ Models Match or Exceed Shallow Variational Quantum Circuits on Vision Benchmarks", "published": "2026-08-07T23:50:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07816", "title": "Preserving Item Semantics for Free: Rethinking Token Initialization in LLM-Based Generative Recommendation", "published": "2026-08-07T23:33:08Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "pretraining-data", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.07814", "title": "Shape Mutating Expert Compression:LorExperts and BTExperts", "published": "2026-08-07T23:23:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07796", "title": "CliniCARE-Bench: Clinical Calibrated Audit of Medical Reasoning in EHR", "published": "2026-08-07T22:39:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07786", "title": "Who Built This Model? Tracing LLM Lineage via Spectral Fingerprints in Weight Space", "published": "2026-08-07T22:18:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07779", "title": "The Capability Ladder: A Curriculum-Modernization Framework for Workforce Readiness in the AI Era", "published": "2026-08-07T21:53:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.10008", "title": "Do LLM Recommenders Know When They're Hallucinating? Auditing Confidence Calibration in Catalog Faithfulness", "published": "2026-08-07T21:41:16Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-amazon", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.07763", "title": "Jako Tako or Fluent? Presenting PoVisLE: A Polish Vision-Language Evaluation", "published": "2026-08-07T21:01:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07742", "title": "BRUCE: Benchmarking Robustness Under Corruption Escalation for Scientific Vision-Language Reasoning", "published": "2026-08-07T20:19:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07737", "title": "The No-Meaning Falsity: The Structural Impossibility of the Arbitrary Sign in Classical Arabic", "published": "2026-08-07T20:01:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07733", "title": "LGNNIC: Acceleration of Large-Scale GNN Training using SmartNICs", "published": "2026-08-07T19:55:45Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07727", "title": "Evaluating Dedicated Monolingual and Joint Multilingual Causal Models for Dravidian Languages", "published": "2026-08-07T19:33:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07700", "title": "Towards Researcher Agents for Knowledge-Graph Question Answering", "published": "2026-08-07T18:42:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07688", "title": "IntelliAudit: Using Large Language Models to Evaluate Audit Controls", "published": "2026-08-07T18:21:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07663", "title": "Keep It Simple: Multi-Key Episodic Memory Retrieval for Ultra-Long Video Understanding", "published": "2026-08-07T18:00:02Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "long-context", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07460", "title": "CreativeInstruct: Scalably Teaching LLMs to Balance Quality, Creativity, and Diversity", "published": "2026-08-07T17:55:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07454", "title": "Strategy-first synthesis planning for complex natural products", "published": "2026-08-07T17:47:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07651", "title": "An Agentic AI Framework Overcomes Fundamental Limitations of Large Language Models for Glaucoma Detection from Fundus Photography", "published": "2026-08-07T17:33:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07440", "title": "Blast Radius", "published": "2026-08-07T17:23:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07439", "title": "An Exploratory Evaluation of LLM-Assisted Rewriting of Moderate-Complexity Financial Sentences for DisCoCat-Based Sentiment Analysis", "published": "2026-08-07T17:23:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07438", "title": "PsychoAgent: An Affect-Sensitive Cognitive Architecture for Conflict-Aware Memory in LLM Agents", "published": "2026-08-07T17:22:29Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "agent-memory", "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07437", "title": "Fisher-R1: Training LLM Agents for Reliable Hypothesis Testing", "published": "2026-08-07T17:22:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07430", "title": "Diffusion LLMs as Targets and Adversaries: Mechanistic Safety Exploits", "published": "2026-08-07T17:17:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11260", "title": "Glance, Scrutinize, and Think: Advancing Video Anomaly Detection from Training-Free to Agentic Reasoning", "published": "2026-08-07T17:07:49Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "multimodal-llm", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07419", "title": "Beyond Post-Hoc Temperature Scaling: Bilevel Optimization for LLM Calibration", "published": "2026-08-07T17:05:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07418", "title": "ResidencyRL: Reinforcement Learning in Simulated Clinical Environments", "published": "2026-08-07T17:04:41Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "multi-agent", "preference-optimization", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.07385", "title": "Omni-modal decomposition autoencoders learn full-stack wearable disentangled representations", "published": "2026-08-07T16:32:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07642", "title": "Contextual Value Alignment via Multilayer Combinatorial Fusion", "published": "2026-08-07T16:22:33Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "multi-agent", "process-reward", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07378", "title": "LSEAD: A Privacy-Preserving LLM-Based Speech Analysis Framework for Early Alzheimer's Disease Screening", "published": "2026-08-07T16:21:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07371", "title": "Trajectory-Relative Hindsight Distillation for Agentic Reinforcement Learning", "published": "2026-08-07T16:12:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07353", "title": "Geo-Spatial Concept Probing of Large Language Models: Abstraction, Compositionality, and Grounding", "published": "2026-08-07T15:46:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07346", "title": "$A^2E$ : An End-to-End Agent Auditing Engine", "published": "2026-08-07T15:44:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07341", "title": "Zero Gap Is Not Restoration: Stratified Per-Question Probability Evaluation and Step-wise Mitigation of Benchmark Contamination", "published": "2026-08-07T15:37:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07333", "title": "When GNNs Fail: Quantifying and Overcoming Temporal Correlation Volatility in Time Series", "published": "2026-08-07T15:28:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07317", "title": "Towards Assurance Closure in AI-Native Large-Scale Agile Software Development", "published": "2026-08-07T15:11:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07282", "title": "Gaze Behavior in Visual World Experiments Can be Modeled With Off-the-shelf Language-Vision Encoders", "published": "2026-08-07T14:42:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07228", "title": "Learning Suffers More Than the Policy Class Under Partial Observability: A Closed-Form Analysis", "published": "2026-08-07T13:42:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07226", "title": "Dual-Node NVIDIA DGX Spark over Tailscale: A Remote-Access Testbed for Distributed LLM Training and Cyber-Threat-Intelligence Fine-Tuning", "published": "2026-08-07T13:41:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07193", "title": "An AI4AI Framework for Visual Token Pruning", "published": "2026-08-07T13:07:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "efficient-inference", "llm-architecture", "multimodal-llm" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.09996", "title": "Energy and Performance Benchmarking of Deep Learning Models for Breast Cancer Detection", "published": "2026-08-07T13:01:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07169", "title": "Agent Memory Distillation: Empowering Small LLM Agents with Hierarchical Teacher Memory", "published": "2026-08-07T12:43:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07161", "title": "Fluid-DiT: Graph-Free Diffusion Transformers for Fluid Flow Simulations Learning", "published": "2026-08-07T12:28:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07148", "title": "A MARL Centered Reference Architecture for Large Language Model Augmentation in Smart Manufacturing", "published": "2026-08-07T12:10:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07147", "title": "DiDPO: Diff-in-Diff Policy Optimization for Coding Agent Training", "published": "2026-08-07T12:07:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07126", "title": "PHOENIX: Fine-Tuned SLM-Powered Autonomous Satellite Lifetime Extension via Predictive Self-Healing and Multi-Agent AI Recovery", "published": "2026-08-07T11:38:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07118", "title": "How Much, Then Where: Credit-Conserving Action-to-Token Allocation for Multi-Turn Agent Reinforcement Learning", "published": "2026-08-07T11:22:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07627", "title": "From Single Chatbots to Governed Agent Ecosystems: An Agentic AI Pattern Catalogue and Orchestration Framework for Mission-Critical Hospital Information Management Systems", "published": "2026-08-07T10:57:30Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "efficient-inference", "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07088", "title": "RoRA: Role-Oriented Regional Allocation for Visual Token Pruning in MLLMs", "published": "2026-08-07T10:39:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07086", "title": "Beyond Isolation: Unlocking Reinforcement Learning Component Synergy for Sample-Efficient Continuous Control", "published": "2026-08-07T10:38:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07079", "title": "LifelongCrossNav: Persistent 3D Semantic Memory for Cross-Floor Multi-Object Navigation", "published": "2026-08-07T10:31:20Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07069", "title": "Invisible to the Machine: Auditing AI Restaurant, Cafe, and Bar Recommendation Against a Complete Market Census", "published": "2026-08-07T10:23:45Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07055", "title": "Teacher Retains Full Tokens, Student Merges Efficiently: TM20K for E-Commerce Sequence Modeling in Ad Recommendation", "published": "2026-08-07T10:04:44Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-bytedance", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.07037", "title": "Accounting Graph Transformer for Short-History Multi-KPI Forecasting in Small Businesses", "published": "2026-08-07T09:47:44Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07036", "title": "CAS2UML: A Handwritten Sketch-to-PlantUML Dataset for Class and Activity Diagrams", "published": "2026-08-07T09:46:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07622", "title": "Controlled Memory Interference in Continual LLM Agents", "published": "2026-08-07T09:46:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07015", "title": "Understand Before Detect: Vision--Language Learning for Omni-Domain Infrared Small Target Detection", "published": "2026-08-07T09:25:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07006", "title": "Does More Retrieved Evidence Help Visual Retrieval-Augmented Generation with Diffusion Language Models?", "published": "2026-08-07T09:20:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07001", "title": "Every Cache Entry Earns Its Place: Global Allocation of Resolution and Coverage for KV Cache Compression", "published": "2026-08-07T09:18:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06997", "title": "Hierarchical Quantization with Domain-Adaptive Sparse Routing for Generative Cross-Domain Recommendation", "published": "2026-08-07T09:15:19Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06993", "title": "Beyond Foundation Models: Dimension-Aware Neural Architecture Search with Small-Data Representation Models for Cryocooler Lifetime Prediction", "published": "2026-08-07T09:11:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07621", "title": "CMU-Drive and V2V-VLA: Cooperative Multi-agent Unified Driving with Reasoning Benchmark and Vehicle-to-Vehicle Vision-Language-Action Models", "published": "2026-08-07T09:00:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06953", "title": "Explicit, Not Longer: What Makes Epistemic Stance Survive Memory Compression", "published": "2026-08-07T08:28:12Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "agent-memory", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06949", "title": "Does Splitting a Triage Decision Across Agents Hide Bias or Help Catch It? A Multi-Agent Simulation Study of LLM-Based Resource Allocation Under Audit Capacity Constraints", "published": "2026-08-07T08:25:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06938", "title": "Debias in Text, Believe Your Eyes: Text-Anchored Cross-Modal Transfer for Visual Counter-Commonsense Reasoning", "published": "2026-08-07T08:10:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06933", "title": "Ask-E: An Environment for Calibrated Question Generation", "published": "2026-08-07T08:06:38Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06931", "title": "Science Edge Evaluation: SEE the Missing Step Toward Real Scientific Discovery", "published": "2026-08-07T08:06:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06928", "title": "From Classification to Recommendation: Empirical Analysis of Audio Embedding Models Application for Content-Based Music Recommendation", "published": "2026-08-07T08:03:29Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06909", "title": "Long-Horizon Agent Trajectory Attribution: A Unified Benchmark and Fine-Grained Annotation Framework", "published": "2026-08-07T07:43:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06908", "title": "Calibrating WEAT Against Anisotropy: ZCA Whitening as a Geometric Pre-Processing Step for Embedding Association Tests", "published": "2026-08-07T07:42:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06901", "title": "Prune Once: Retraining-Free Task-Agnostic Pruning for Vision-Language Models", "published": "2026-08-07T07:36:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "model-compression", "multimodal-llm" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.06865", "title": "Multi-Agent Forensic Reasoning for Generalizable Deepfake Video Detection", "published": "2026-08-07T06:44:14Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06849", "title": "Autonomy-of-Heads: Data-Free Sparse Attention from Frozen Query-Key Geometry", "published": "2026-08-07T06:18:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "efficient-inference", "long-context", "model-compression" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B07", "plan_reason": "LLM 架构、长上下文、KV cache 与评测基础设施 — completed" }, { "arxiv_id": "2608.06834", "title": "Graph Machine: Exploring Edge Mechanisms as an Inductive Bias", "published": "2026-08-07T05:48:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06825", "title": "Multiscale Reward Hedging from Correct Demonstrations", "published": "2026-08-07T05:36:22Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06811", "title": "Coupling Planning with Episodic Memory in LLM Agents for Software Issue Resolution", "published": "2026-08-07T05:05:05Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agent-planning", "software-agent" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B11", "plan_reason": "Agent 记忆、工具规划与自进化系统 — completed" }, { "arxiv_id": "2608.06802", "title": "Simple-OPD: Demystifying Warm-up for On-policy Distillation", "published": "2026-08-07T04:47:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06795", "title": "LoRAScan: Detecting Backdoor Prompts in Low-Rank Adapters for Large Language Models via Down-Projection Activation Spikes", "published": "2026-08-07T04:36:24Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06792", "title": "Progressive Alignment of Recommender Foundation Model through Multi-Phase Post-Training", "published": "2026-08-07T04:30:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.06791", "title": "HLSmith: An Expert-Guided Agentic Framework for C/C++-to-HLS Translation", "published": "2026-08-07T04:27:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06790", "title": "AgentChaos: Chaos Engineering for Agent Systems via Programmatic Fault Injection", "published": "2026-08-07T04:24:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06778", "title": "Retrieval-Constrained Policy Optimization for Attack Technique Extraction from Cyber Threat Intelligence", "published": "2026-08-07T03:52:33Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-rl", "multi-agent", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06769", "title": "GraphVerse: A Comprehensive Visual Graph Reasoning Benchmark for Multimodal Large Language Models", "published": "2026-08-07T03:43:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06763", "title": "CubicQuant: Parametric Non-Uniform Codebooks for High-Throughput LLM Inference with 1-8-Bit Weights", "published": "2026-08-07T03:36:07Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06758", "title": "Stockmark-Nemotron-3-Nano-Omni-JapanDocReader: Structured Document Parsing via Capability Injection and Forgetting Control", "published": "2026-08-07T03:24:59Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "multimodal-llm", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06756", "title": "Capek 0.5: An Execution-Centric Vision-Language Model for Embodied Intelligence", "published": "2026-08-07T03:24:25Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06750", "title": "Progressive Content Refinement with Decaying Reward Joint LinUCB", "published": "2026-08-07T03:17:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06735", "title": "IB-RL: Isolated Bilateral Reinforcement Learning for Strategic Dialogue Agents", "published": "2026-08-07T02:59:53Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06723", "title": "Multi-Level Modeling of Large Language Model Inference Latency and Energy via Hybrid Analytical--Machine-Learning Predictors", "published": "2026-08-07T02:37:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06701", "title": "Online Monitoring and Corrective Steering of Programming Agents", "published": "2026-08-07T01:54:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06699", "title": "AgentPatch: Coarse-to-Fine Weak-Task Repair for Merging Agentic Multimodal Large Language Models", "published": "2026-08-07T01:52:45Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multimodal-llm", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06694", "title": "A Multi-Agent Framework for Automated Coarse-Grained Molecular Dynamics of Polymers", "published": "2026-08-07T01:47:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06690", "title": "Policy-Masked Private Experts: Auditable and Reversible Capability Access Control in Sparse MoE Models", "published": "2026-08-07T01:39:36Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06672", "title": "TA-RAG: Tone Awareness as a Design Imperative for Retrieval-Augmented Generation", "published": "2026-08-07T00:37:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06668", "title": "Vehicle routing problem using deep reinforcement learning - A case study about truck planning in the industry", "published": "2026-08-07T00:25:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06663", "title": "The Horizon Gap: Planning, Memory, Execution, Training, and Evaluation for Long-Horizon LLM Agents", "published": "2026-08-07T00:19:48Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning", "long-context", "process-reward", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06651", "title": "CyberLLM: A Multi-Agent LLM Framework for Autonomous Detection and Guarded Response in Automotive Cybersecurity", "published": "2026-08-06T23:39:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06630", "title": "The Sparsity Whisperer", "published": "2026-08-06T22:37:26Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06628", "title": "Retrofitting Linear Attention into Diffusion Language Models", "published": "2026-08-06T22:34:20Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14689", "title": "A Reproducibility Study of Partial Residual Ablations in Pre-LN Transformers", "published": "2026-08-06T21:29:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06595", "title": "Flowing Through States: Neural ODE Regularization for Reinforcement Learning", "published": "2026-08-06T21:11:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06564", "title": "Which Decisions Low-Bit Quantization Breaks, and How to Predict Them", "published": "2026-08-06T20:22:21Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06526", "title": "GRASP: Reinforcing Language Model Anonymizers with Group Relative Policy Optimization", "published": "2026-08-06T19:12:57Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "on-policy-distillation", "opd", "preference-optimization", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.06503", "title": "Toward Reliable Context Compression for Long-Horizon Agents: An Empirical Study of Execution Instability", "published": "2026-08-06T18:42:02Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06501", "title": "Can MLLMs Decode the Creative Leap? Introducing C4 for Cross-Concept Understanding", "published": "2026-08-06T18:38:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06486", "title": "Beyond Attention: Signed Integrated Gradients Attribution in a BiomeGPT-Style Microbiome Transformer", "published": "2026-08-06T18:26:05Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06485", "title": "Do AI Personas Grow? Analyzing and Benchmarking Personality Evolution in LLM Agents After Life Events", "published": "2026-08-06T18:25:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06474", "title": "WebGrader: Training LLMs for Web Development with Self-Evolving Programmatic Grader", "published": "2026-08-06T18:06:12Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-planning", "agentic-rl", "web-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.06471", "title": "CyberForge: Verified Vulnerability Injection at Repository Level for Cybersecurity Agent Training", "published": "2026-08-06T18:03:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06377", "title": "Learning When to Trust via Selective Context Preference Optimization", "published": "2026-08-06T17:59:58Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06370", "title": "The Bitter Lesson of Tool Calling", "published": "2026-08-06T17:58:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06366", "title": "Tracing the Heart: An Evidence-Linked Pipeline for Heart-Failure Feature Engineering", "published": "2026-08-06T17:57:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06362", "title": "AV-AIVAT: 74x Cheaper Agent Evaluation with Certified Anytime-Valid Stopping in Imperfect-Information Games", "published": "2026-08-06T17:57:11Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06312", "title": "Benchmarking and Enhancing LLMs for Rule-Intensive Review of National Standard Documents", "published": "2026-08-06T17:27:23Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06305", "title": "Beyond Top-K: Replacing Black-Box Retrieval with Interpretable Agentic Operations", "published": "2026-08-06T17:23:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06246", "title": "A Six-Dimensional Taxonomy of Post-Training Adaptation Techniques with Applications in AI Governance", "published": "2026-08-06T16:32:26Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06241", "title": "Timestep-Conditioned Transformers for Global Weather Forecasting", "published": "2026-08-06T16:27:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06223", "title": "TS-RAG: Retrieval Augmented Generation for Time Series Forecasting", "published": "2026-08-06T16:12:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06179", "title": "SAGA: Score-Weighted Adaptive Generation Alignment for Low-Resource Nordic Language Models", "published": "2026-08-06T15:41:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06171", "title": "Routing Is Least Learnable Where It Is Most Valuable: Bounds on Representation Routing for Web Agents", "published": "2026-08-06T15:37:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06161", "title": "iARCS: Iterative Agentic RL for Controllable 3D Scene Generation", "published": "2026-08-06T15:30:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06122", "title": "Is Self-Pretraining really useful to improve diagnosis in medical Time Series?", "published": "2026-08-06T14:53:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06113", "title": "DCAS: Decoupling CLI Agent Scaffolding to Internalize Planning across Scaffolds", "published": "2026-08-06T14:47:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06112", "title": "From Siloed Algorithms to Compliance-First Agentic Platforms: A Multi-Layered Architecture for Hospital AI Systems", "published": "2026-08-06T14:44:31Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-architecture", "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06111", "title": "Beyond Sequence Order: Syntax-Informed Positional Embeddings for Transformers", "published": "2026-08-06T14:44:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06068", "title": "Cleo: A Transparent and Controllable Chatbot for Conversational Commerce", "published": "2026-08-06T14:12:14Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recommendation-ranking", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.06060", "title": "Learning from Failures: Retrieval-Centric CoT via Hard Negatives for Unified Multimodal Retrieval", "published": "2026-08-06T14:04:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06041", "title": "LangChoiceBench: Measuring and Explaining Programming-Language Choice in LLMs", "published": "2026-08-06T13:52:16Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06027", "title": "FormBharo: Designing and Evaluating a Voice Agent for Conversational Form Filling in Rural India", "published": "2026-08-06T13:34:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06022", "title": "EpiBench: Can LLMs Understand Epitopes for Antibody Drug Discovery?", "published": "2026-08-06T13:29:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06015", "title": "ProDVI: Programmatic Dynamics Priors for Value Network Initialization", "published": "2026-08-06T13:19:29Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05980", "title": "How Far Do Simple Transformations Translate Across Text Embedding Models?", "published": "2026-08-06T12:57:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05959", "title": "AgentExecutor: Partial Code Execution via Agentic Context Generation", "published": "2026-08-06T12:33:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05954", "title": "Training a Conditioned Video Game Agent on a VLM Annotated Dataset", "published": "2026-08-06T12:25:54Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05949", "title": "VLMs for Videogame Data Annotation", "published": "2026-08-06T12:20:53Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05906", "title": "Causal Episodic Memory for Feedback-Driven Agent Repair", "published": "2026-08-06T11:34:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18142", "title": "Efficient Adaptation of LLMs for Hate Speech Detection in Low-Resource Languages: A Comparative Study on Roman Urdu", "published": "2026-08-06T11:28:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05891", "title": "AppDeltaWorld: Transition-Grounded Delta Code World Model for Mobile GUI Agents", "published": "2026-08-06T11:15:41Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05850", "title": "MameLoshnLM: Yiddish Language Model and Evaluation Benchmark", "published": "2026-08-06T10:24:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05832", "title": "Enhancing Social Intelligence in LLMs with Hierarchical Reasoning and Utterance-Level Goal Rewarding", "published": "2026-08-06T10:00:50Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-rl", "multi-agent", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05825", "title": "MoCA: Implicit Social Context Analysis", "published": "2026-08-06T09:54:24Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05822", "title": "Agent-Based Test Assertion Generation via Diverse Perspective Aggregation", "published": "2026-08-06T09:51:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05817", "title": "M$^3$R-Bench: A Unified Benchmark for Evidence-Grounded Multimodal Metaphor Understanding", "published": "2026-08-06T09:48:36Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05802", "title": "On-Policy Delta Distillation for Multilingual Math Reasoning", "published": "2026-08-06T09:37:49Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation", "post-training" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B08", "plan_reason": "OPD 与多教师/过程蒸馏 — completed" }, { "arxiv_id": "2608.05792", "title": "When Agentic AI Meets Integrated Sensing and Communication", "published": "2026-08-06T09:26:49Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-planning", "agentic-rl", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.05791", "title": "A Two-Tier Perspective on Inference-Time Parallelism in Multi-Agent LLM Systems", "published": "2026-08-06T09:26:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05790", "title": "ChainClaw: A Layered Agent Framework for Reliable On-Chain Execution", "published": "2026-08-06T09:25:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05774", "title": "SR-JEPA: Learning Predictive Latent State in 3D Scenes", "published": "2026-08-06T09:09:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05745", "title": "UniVVT: A Unified End-to-End Framework for High-Fidelity Video Virtual Try-on", "published": "2026-08-06T08:29:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05729", "title": "Unified Agent: Managing Interactions across Devices", "published": "2026-08-06T08:14:33Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "multimodal-llm", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05703", "title": "StreamArena: Toward Continuous, Interactive, and Long-Horizon Agentic Streaming Video Understanding", "published": "2026-08-06T07:46:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05695", "title": "DreamGuard: Efficient Runtime Guardrail for LLM Agents via Risk-Aware World Model", "published": "2026-08-06T07:37:49Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05683", "title": "DistMedVL: Distributional Vision-Language Alignment for Uncertainty-Aware Medical Image Segmentation", "published": "2026-08-06T07:17:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05655", "title": "Is Personalized Modality Weighting Actually Personalized? A Controlled Audit of Per-User Weighting Claims in Multimodal Recommenders", "published": "2026-08-06T06:52:38Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.05651", "title": "Relay, Don't Route: Adaptive Population Handoff for Cost-Efficient LLM-Driven Evolution", "published": "2026-08-06T06:48:08Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "efficient-inference", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05648", "title": "Vorch-IR: Long-Form Unified Multimodal Identity Replacement Video Generation", "published": "2026-08-06T06:46:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05643", "title": "Refining Over Resampling: Test-Time Self-Correction for LLM Reasoning", "published": "2026-08-06T06:38:37Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06434", "title": "Fast and Accurate: An Adaptive VLA Inference Framework through Environment-aware Model Selection", "published": "2026-08-06T06:32:27Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-architecture", "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05631", "title": "ChronoVision: Temporal Reasoning via Latent State Reconstruction", "published": "2026-08-06T05:58:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05604", "title": "SkillZip: Contract-Preserving Graph Compression for Scalable Agent Skill Libraries", "published": "2026-08-06T05:03:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05600", "title": "LC-GRPO: Bridging Train-Inference Gap for Flow-Based GRPO with Langevin Correction", "published": "2026-08-06T04:47:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05592", "title": "Beyond Frame Selection: Rethinking Long-Video Understanding with MLLMs", "published": "2026-08-06T04:27:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05588", "title": "Search-Aided Joint Agent-Environment Reinforcement Learning for Robust Lifelong Multi-Agent Path Finding with Rotations", "published": "2026-08-06T04:17:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05539", "title": "OmniMech: All-in-one Multimodal Mechanical Benchmark for 3D Reconstruction", "published": "2026-08-06T02:36:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20405", "title": "ARGUS: Theory-of-Mind Guided Argument Generation with Strategy-Aware Planning and Knowledge Grounding", "published": "2026-08-06T02:26:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05521", "title": "Reasoning from Traces: Divergence-Guided Agentic Repair of WebAssembly Discrepancies", "published": "2026-08-06T01:52:20Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05519", "title": "EcoAgent-Bench: Evaluating Economic Decision-Making in Budget-Constrained LLM Agents", "published": "2026-08-06T01:47:47Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "reward-model", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05499", "title": "APQF: Agentic Profiling-Guided Structured Pruning and Mixed-Precision Quantization with Adaptive Fine-Tuning", "published": "2026-08-06T01:09:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05493", "title": "Learning Context-Free Grammars for Grammar-Constrained Decoding via Declarative Agentic Programming with Guarantees", "published": "2026-08-06T00:44:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05475", "title": "KV-Skill: Forging Expertise in the Model's Native Language", "published": "2026-08-05T23:46:56Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "post-training", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05472", "title": "Matrix Zonotopic Attention: A Context-Adaptive Value Projection for Set Transformers", "published": "2026-08-05T23:43:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05409", "title": "Mood Matters: How Syntactic Sensitivity Undermines Safety Alignment", "published": "2026-08-05T21:05:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07593", "title": "Weather- and Location-Aware Agentic Dining Recommendation: Leveraging LLM World Knowledge for Region-Sensitive Contextual Reasoning", "published": "2026-08-05T20:55:24Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-google", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.05402", "title": "Robustness and User-Perceived Value of Popularity Calibration in Music Recommendation: A User Study", "published": "2026-08-05T20:44:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05375", "title": "DoctorAgents: an agentic framework to iteratively refine AutoML pipeline for small clinical temporal data", "published": "2026-08-05T19:52:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05359", "title": "CASCADE: An Agentic Regulatory Network Framework for Patient-Data-Validated Downstream Perturbation Prediction", "published": "2026-08-05T19:30:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05346", "title": "Multi-Agent Reinforcement Learning for Online Traffic Scheduling in Time-Sensitive Application", "published": "2026-08-05T19:06:42Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06425", "title": "NTDH: Complex Reasoning for Comprehensive Affective Analysis", "published": "2026-08-05T18:59:43Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05341", "title": "Positive-Unlabeled Preference Optimization For Chest X-ray Report Generation", "published": "2026-08-05T18:59:23Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06424", "title": "Multi Codec Discrete Diffusion Model for Text Guided Speech Inpainting and Editing", "published": "2026-08-05T18:57:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05340", "title": "Multi-Agent Transformer for Queue-Level XR Traffic Scheduling in TSN Networks", "published": "2026-08-05T18:57:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05303", "title": "EdgeXpert: An Edge Device for Memory-Efficient LLM Inference with Mixture-of-Experts and Speculative Decoding", "published": "2026-08-05T18:03:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06417", "title": "Latent Fact-Checking: Detecting Misinformation through Activation Engineering", "published": "2026-08-05T18:00:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05144", "title": "Argus: A General-Purpose Agentic Reasoning Runtime for Long-Horizon Tasks", "published": "2026-08-05T17:58:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05266", "title": "Agentic self-driving microscopy benchmarks support qualification but do not necessarily generalize to unseen tasks", "published": "2026-08-05T17:58:53Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05141", "title": "OctoLong: Mid-Training On Cross-Repository Code Contexts Enhances Long-Context Modeling", "published": "2026-08-05T17:58:15Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05139", "title": "Toward Skill-Native LLMs: Skill Entropy for Benchmarking and Training Long-Horizon Reasoning", "published": "2026-08-05T17:57:16Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05137", "title": "SmartMage: Dynamic Modality Orchestration for 3D Scene Understanding", "published": "2026-08-05T17:56:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05131", "title": "OPD-V: Visual On-Policy Self-Distillation with Modality Balance", "published": "2026-08-05T17:53:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05126", "title": "Spoken Function Calling: A New Perspective on Spoken Language Understanding for Large Audio Language Models", "published": "2026-08-05T17:50:31Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05124", "title": "Chained Recursive Language Models for Multi-Iteration Reasoning", "published": "2026-08-05T17:50:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05116", "title": "Characterizing Visual Accessibility Issues in AI Developer Tools: An Empirical Study", "published": "2026-08-05T17:46:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05111", "title": "Reward Structure Shapes the Interaction Between Episodic Exploration and Neural Memory in Reinforcement Learning", "published": "2026-08-05T17:44:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05107", "title": "CoPlan: A Trustworthy Co-Intelligence Interface for Care Planning through Role-Based Contestable Argument Graphs", "published": "2026-08-05T17:43:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05102", "title": "ABSeeker: Training Long-Horizon Search Agents via Answer-Backtracked Credit Assignment", "published": "2026-08-05T17:41:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05086", "title": "Item Response Theory for AI Safety", "published": "2026-08-05T17:25:27Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05050", "title": "The Effect of Perceived Race and Gender on Police Language Use: Experimental Evidence from VR Simulations", "published": "2026-08-05T16:59:49Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05255", "title": "An Emerging Retail Portfolio Management Application: Personalized, Tax-Aware Reinforcement Learning with Natural Language Goals", "published": "2026-08-05T16:20:11Z", "tracks": [ "post-training", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-rl", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05000", "title": "Towards Physics of Multimodal Pretraining: Knowledge Flow, Modality Synergy, Early Unification, and Recipes", "published": "2026-08-05T16:09:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "multimodal-llm", "pretraining-data" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B07", "plan_reason": "LLM 架构、长上下文、KV cache 与评测基础设施 — completed" }, { "arxiv_id": "2608.04949", "title": "UG-UMRE: Uncertainty-Guided Modality Augmentation and Distributional Calibration for Unified Multimodal Relation Extraction", "published": "2026-08-05T15:18:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05245", "title": "Search2Skill: Skill Distillation Beyond Knowledge Boundaries Via Rubric-Based Reinforcement Learning", "published": "2026-08-05T15:09:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04939", "title": "Reading Between the Frames: Interpreting Implicit and Non-literal Meaning in Social Media Videos", "published": "2026-08-05T15:06:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04904", "title": "Strengthening Target-Language Features: SAE-Based Steering for Multilingual Inference", "published": "2026-08-05T14:32:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04879", "title": "Training Crossroads for Recurrent Vision Transformers: Recurrence, Neural ODEs, and Deep Supervision", "published": "2026-08-05T14:06:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04872", "title": "A-SR: Self-Evolving Agentic LLMs for Symbolic Regression via Hierarchical Coordination", "published": "2026-08-05T14:01:10Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05235", "title": "From Trajectories to Evidence: Auditable Experimental Records for Industrial Research Agents", "published": "2026-08-05T13:37:38Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04830", "title": "ContextWeave: A Real-World Workflow Benchmark", "published": "2026-08-05T13:31:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04828", "title": "Skill-Use: Can LLMs Actually Use Skills in Agentic Harnesses?", "published": "2026-08-05T13:29:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04807", "title": "WatchLens: A Configurable Platform for Online Video Recommendation Experiments", "published": "2026-08-05T13:13:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.04804", "title": "Scrouting: Cost-Aware Routing of Coding Agents by Scouting the Repository First", "published": "2026-08-05T13:11:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04794", "title": "Privileged, but Biased: How PI-Conditioned Teachers Break Self-Distillation", "published": "2026-08-05T12:59:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04768", "title": "Embedding Large Language Models into Flow Controls: An Agentic Framework for Adaptive and Trustworthy Automated Cooking", "published": "2026-08-05T12:33:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04759", "title": "Trace, Verify, and Correct: A Training-Free Framework for Spatial Reasoning in Multimodal LLMs", "published": "2026-08-05T12:26:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04753", "title": "Attention, Anomalies! Handling Attention Layers in Unsupervised Federated Outlier Detection", "published": "2026-08-05T12:23:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04750", "title": "Simile Understanding in Text-to-Image Models: An Evaluation Framework", "published": "2026-08-05T12:18:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04726", "title": "When Prompts Become Pixels: Prompt-Region Grounding for Multimodal Reasoning", "published": "2026-08-05T11:46:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05223", "title": "Towards a Risk Assessment of Malicious Skill Files in Coding Agents", "published": "2026-08-05T11:33:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04714", "title": "What We Observe as LLM Behavior Can Be a Side-effect of Inference Backend", "published": "2026-08-05T11:28:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04698", "title": "Teaching MLLMs to Say No: Generalized Referring Expression Comprehension via Refusal Calibrated GRPO", "published": "2026-08-05T11:07:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04682", "title": "Active-SWE: Benchmarking Coding Agents for Proactive Bug Fixing without Issue Reports", "published": "2026-08-05T10:50:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04678", "title": "Kathleen Writes: Autoregressive Generation and Data Scaling Without Attention", "published": "2026-08-05T10:44:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "pretraining-data", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.04677", "title": "Diverse and Plausible Algorithmic Recourse via Tractable Recourse Distributions", "published": "2026-08-05T10:40:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04663", "title": "Calibrating Artificial Guilt: Neurally Grounded Reward Shaping for Prosocial Multi-Agent Reinforcement Learning", "published": "2026-08-05T10:21:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04661", "title": "An Exploratory Study of Agent Plans for Agentic AI Coding Tools in Open-Source Software", "published": "2026-08-05T10:15:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04646", "title": "Evaluating Theory of Mind in Reasoning Models: Robustness over Reasoning", "published": "2026-08-05T10:05:53Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "reward-model", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.04589", "title": "The First EgoCross Challenge at EgoVis 2026: Cross-Domain Egocentric Video Question Answering", "published": "2026-08-05T08:51:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04588", "title": "EASy: Towards Efficient LLM-Based Agentic System", "published": "2026-08-05T08:50:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04586", "title": "Breaking the Curse of Multilinguality in Many-to-Many Speech-to-Text Translation via a Resource-Aware Mixture of Speech Encoders", "published": "2026-08-05T08:49:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04576", "title": "Causal Evidence Extraction and Triangulation in Crisis Reports using Large Language Models: A ReliefWeb-based Study", "published": "2026-08-05T08:07:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04574", "title": "When Memory Lies: An Empirical Study of Spatial Memory Staleness in VLM Agents", "published": "2026-08-05T08:04:26Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04569", "title": "Relevant but Incomplete: Referential Dangling as a Paradigm-Level Failure Mode in Hard Prompt Compression", "published": "2026-08-05T07:58:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04565", "title": "Breadcrumbing Search Agents", "published": "2026-08-05T07:57:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05207", "title": "When Do Corrective Features Help? An Agent for Corrective Feature Discovery on Black-Box Forecasters", "published": "2026-08-05T07:48:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04533", "title": "EgoAfford: Task-Oriented Affordance Grounding via Egocentric Referring Segmentation", "published": "2026-08-05T07:08:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04515", "title": "CARVE: Cross-Slice Anisotropic Reallocation of Visual Evidence for Efficient 3D Medical Volume Understanding", "published": "2026-08-05T06:49:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04514", "title": "RESPClinBench: Benchmarking Multimodal Clinical Decision-Making and Longitudinal Disease Management in Respiratory Specialty Care", "published": "2026-08-05T06:49:30Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multimodal-llm" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04505", "title": "K-EXAONE 2.0 Technical Report", "published": "2026-08-05T06:41:48Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "long-context", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04488", "title": "Energy- and Memory-Efficient PEFT Methods for Personalized On-Device SLMs on Consumer GPUs", "published": "2026-08-05T06:20:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04477", "title": "DeepInvert: Semi-Supervised Embedding Inversion Against Obfuscated Language Models", "published": "2026-08-05T06:02:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04460", "title": "Tropical Algebraic Geometry for Neuronal Representations: An Arakelov-Green Measure Based Descriptor for Graph Learning", "published": "2026-08-05T05:34:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04459", "title": "AdaptAgent: A Multi-agent, Domain-Guided Reasoning Framework for Code Adaptation", "published": "2026-08-05T05:34:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16928", "title": "Benchmarking Classical and Transformer-Based Models for Document Sensitivity Classification", "published": "2026-08-05T05:00:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14685", "title": "Rethinking Reverse KL as Adaptive Entropy Distillation", "published": "2026-08-05T04:27:37Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04428", "title": "Deltoris: Enabling Real-time VLA Inference in Embodied AI via Bit-level Sparsity and Speculative Inference", "published": "2026-08-05T04:17:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14684", "title": "Mitigating Rubric Interference in LLM Judges via On-Policy Self-Distillation", "published": "2026-08-05T03:35:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04405", "title": "Training-Free Hashing-Based Attention via Binary Principal Components", "published": "2026-08-05T03:19:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04390", "title": "EdgeLM: Edge Demonstrations for Language Models' Table Understanding", "published": "2026-08-05T02:45:24Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04366", "title": "Combating Knowledge Corruption in Agent Systems: A Byzantine-Tolerant Secure Collaborative RAG Framework", "published": "2026-08-05T02:13:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04330", "title": "Right Reset: Chunking by Prefix Removal", "published": "2026-08-05T01:17:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04317", "title": "Trident : How to Break Deep Reinforcement Learning Cyber Defenses (Agentic)", "published": "2026-08-05T00:54:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14681", "title": "Automatic or Controlled? Repetition Priming Reveals Divergent Processing in Base LLMs, Instruct LLMs, and Humans", "published": "2026-08-05T00:52:03Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04311", "title": "Pun Intended: Multi-Agent Translation of Wordplay with Contrastive Learning and Phonetic-Semantic Embeddings", "published": "2026-08-05T00:40:42Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04299", "title": "Searching for Sound-Meaning Collisions: Graph-Based Affordance Retrieval and Multi-Evaluator Ranking for Pun Translation at CLEF 2026 JOKER Task 2", "published": "2026-08-05T00:06:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04286", "title": "Eliciting Intrinsic Hallucinations in LLMs via Semantically Equivalent Adversarial Attacks", "published": "2026-08-04T23:25:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04278", "title": "EA-Graph: Artifact-Anchored Verification Memory for Coding Agents under Upstream Drift", "published": "2026-08-04T23:13:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04270", "title": "CURATE: Leveraging LLM Agents to Compose, Catalog, and Deploy Reproducible Workflows", "published": "2026-08-04T23:01:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04268", "title": "The Fairness Collapse Phenomenon: Bias Amplification in Language Models Trained on Synthetic Data", "published": "2026-08-04T22:56:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14680", "title": "When Agentic Executions Fail: Detecting and Localizing Runtime Faults from Telemetry", "published": "2026-08-04T20:47:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04213", "title": "Attention-Only White-Box Transformer via LeJEPA-Based Self-Supervised Pretraining", "published": "2026-08-04T20:26:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04195", "title": "SONAR: Task-Aware Code Summary Evaluation for LLM Consumers Without References", "published": "2026-08-04T19:54:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04186", "title": "Large Language Models for Low-Resource Languages: A Conceptual Framework for an Electronic Explanatory Dictionary of the Tajik Language", "published": "2026-08-04T19:43:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04148", "title": "AgentForge: An Immersive Role-Playing Platform for Learning Agentic Software Engineering", "published": "2026-08-04T18:58:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04095", "title": "FinPerMA: A Theory-Informed, Event-Grounded Personalized-Memory Benchmark for LLM Agents", "published": "2026-08-04T18:00:04Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04009", "title": "SocietyBench: Forecasting Counterfactual Social-World Evolution", "published": "2026-08-04T17:59:56Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "agent-memory", "post-training", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04007", "title": "TurnSight: Turn-Level Hindsight Self-Distillation for Tool-Integrated Reasoning", "published": "2026-08-04T17:59:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04003", "title": "PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents", "published": "2026-08-04T17:58:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03999", "title": "Agogic: Performance-Timed Music Tokens for LLM-Native Text-to-Symbolic-Music Generation", "published": "2026-08-04T17:56:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05013", "title": "OneDayAgent: Towards a Long-Horizon Harness for Autonomous Agents", "published": "2026-08-04T17:55:41Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03970", "title": "Should We Type or Talk to LLM Agents? A Comprehensive Study of Voice and Keyboard Input Perturbations", "published": "2026-08-04T17:38:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03930", "title": "Logic Before Language: Pre-pretraining on Formal Derivations Fosters Skill Acquisition and Compressibility", "published": "2026-08-04T17:02:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "long-context", "model-compression", "pretraining-data" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.03929", "title": "Latent Reward Registers for Diffusion Preference Alignment", "published": "2026-08-04T17:00:52Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03927", "title": "A Physics-Flavored Transformer Network for Parametrizing Contraction Dynamics of Engineered Skeletal Muscle Tissues", "published": "2026-08-04T16:59:30Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03913", "title": "Sparse Weight Decomposition for Efficient Circuit Extraction", "published": "2026-08-04T16:40:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03899", "title": "ATLAS: Learning to Recommend Across Unseen Domains", "published": "2026-08-04T16:29:02Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-amazon", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.03898", "title": "ANNOTARES: A Dataset for Extracting Logical Structures from German Statutory Texts", "published": "2026-08-04T16:28:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03884", "title": "BanglaWild: An In-the-Wild Bengali Scene Text Recognition Benchmark for OCR and Vision-Language Models", "published": "2026-08-04T16:20:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03882", "title": "MultiGlobeQA: A Multilingual and Globally Diverse Benchmark for Geospatial Reasoning", "published": "2026-08-04T16:18:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03875", "title": "Enhancing VLM Reward Models Through Structure-Aware Fine-Tuning", "published": "2026-08-04T16:15:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03855", "title": "Bi-semantic Chemical Embedder for Joint Representation Learning of SMILES and Natural Language", "published": "2026-08-04T15:57:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "long-context", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.03852", "title": "FedCritic-MIMO: Communication-Efficient Serverless Federated Critic Learning for Massive-MIMO Resource Control in Open and Disaggregated 6G RANs", "published": "2026-08-04T15:56:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03836", "title": "Resume Means Resume: A Machine-Checked Conformance Contract for Checkpoint, Interrupt, and Resume Semantics in Workflow Persistence Layers", "published": "2026-08-04T15:45:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03803", "title": "M-GATE: Multilingual Grammar, Accuracy in Translation, and Efficiency Benchmark for Large Language Models", "published": "2026-08-04T15:16:05Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03800", "title": "Autoreflection: How Agentic Strange Loops Turn Human Culture into AI Infrastructure", "published": "2026-08-04T15:14:50Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "multi-agent", "web-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.03796", "title": "Efficient Knowledge Distillation for LLMs: Offline Top-K Logits and a Fused Chunked KL Loss", "published": "2026-08-04T15:11:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04066", "title": "The LLM Proposes, the Executive Disposes: A Self-Verifying Agent Instrument that Dissociates Commitment Drift from Binding Drift in Long-Horizon Agents", "published": "2026-08-04T15:10:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03722", "title": "When Outputs Disperse, Does Epistemic Revision Follow? A Black-Box Diagnostic for Machine Collectives", "published": "2026-08-04T14:19:59Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "llm-recommendation", "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.03692", "title": "SITA: Semantic Interest Tokens for Target-Aware Compression in Long-Sequence Recommendation", "published": "2026-08-04T13:59:44Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03689", "title": "LiveEvalBench: Toward Open-World Evaluation for Web Generation", "published": "2026-08-04T13:57:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03673", "title": "CausalOPD: First-Wrong-Step Supervision for Distilling Causal Chain Reasoning", "published": "2026-08-04T13:49:12Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation", "opd" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B08", "plan_reason": "OPD 与多教师/过程蒸馏 — completed" }, { "arxiv_id": "2608.03659", "title": "How Closely Do LLM Reviews Align with Human Peer Review?", "published": "2026-08-04T13:39:36Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "llm-recommendation", "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.03644", "title": "Is Inter-Seed Cross-Play Enough? Evaluating the Robustness of Zero-Shot Coordination Algorithms to Implementation Details", "published": "2026-08-04T13:29:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03629", "title": "Cross-Layer Interaction under Weight-Space Ablation: A Closed-Form Attention Jacobian Bound and a Test on a Real Pretrained Model", "published": "2026-08-04T13:14:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03620", "title": "A Theory of Conditional Collapse under Low-Rank Weight-Space Ablations: I. The Single-Block Theory and Synthetic Validation", "published": "2026-08-04T13:10:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03610", "title": "Language-Specialized Multi-Teacher On-Policy Distillation for Multilingual LLM-Based ASR", "published": "2026-08-04T13:02:47Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03606", "title": "Learning Clinical-Trial Strategy: Offline Policy Training for Decision Agents", "published": "2026-08-04T12:58:17Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03598", "title": "From Bug Reports to Browser-Executable Procedures: An LLM-Driven Agent for Web GUI Bug Reproduction", "published": "2026-08-04T12:50:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03588", "title": "GenOS: Compositional Certificates for Semantic Robustness in AI Code Generation", "published": "2026-08-04T12:42:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03579", "title": "Pin Once, Swap Light: Subspace-Aligned Centroid-Residual Training for Efficient Ultra-LoRA Serving", "published": "2026-08-04T12:34:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03573", "title": "SFT Conflicts, RL Coexists: A Theoretical and Empirical Analysis of Multi-Task Learning for LLMs", "published": "2026-08-04T12:32:26Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03558", "title": "EffiHolmes: Differential Profiling-Guided Repository Level Time Inefficiency Fix Localization", "published": "2026-08-04T12:26:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03545", "title": "Hi-TTRL: Regulating Consensus with Hints for Test-Time Reinforcement Learning", "published": "2026-08-04T12:20:17Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "reward-model", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.03527", "title": "Training Documents Reranker with Search Rubrics for Deep Research Agent", "published": "2026-08-04T12:11:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03502", "title": "Hybrid LLM-Augmented Reinforcement Learning Agents for Complex Sequential Decision Tasks", "published": "2026-08-04T11:44:07Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-planning", "agentic-rl", "llm-architecture", "llm-rl", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.04056", "title": "Learning Sexism Detection Using Multi-Agent Perspectivist Preference Optimization", "published": "2026-08-04T11:35:30Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "preference-optimization", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03483", "title": "Continue or Replan? Bernoulli-Continuation Policy Learning for Adaptive Horizon Execution", "published": "2026-08-04T11:21:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03480", "title": "Efficient Multilingual Neural Machine Translation via Corpus-Driven Vocabulary Pruning: An English-Arabic Case Study", "published": "2026-08-04T11:17:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03475", "title": "Adaptive Modality Reliability Diagnosis and Restoration for Robust Multimodal Intent Recognition", "published": "2026-08-04T11:14:13Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03468", "title": "ToolLIFT: Lifting Tool-Specific Trajectories into Function-Level Graphs for Generalizable Tool Planning", "published": "2026-08-04T11:02:45Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-planning", "agentic-rl", "tool-agent" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B11", "plan_reason": "Agent 记忆、工具规划与自进化系统 — completed" }, { "arxiv_id": "2608.03452", "title": "Probing Character-level Transformers for the Spanish L-shaped Morphome", "published": "2026-08-04T10:48:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03450", "title": "Balancing Efficiency and Efficacy: Training-Free Attention-Guided Switching Between Explicit and Latent Thoughts for MLLMs", "published": "2026-08-04T10:46:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03420", "title": "Towards Improving Sequential Decision-Making in LLM Agents via Experience Memory", "published": "2026-08-04T10:12:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03392", "title": "Self-Evolving Coding Agents", "published": "2026-08-04T09:43:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03391", "title": "TimeRLM: Recursive Language Models Enable Precise Anomaly Localization in Long-Context Time-Series", "published": "2026-08-04T09:40:45Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "long-context", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04048", "title": "Recurrent Residual Quantization: A Progressive Multi-Precision Representation for LLMs", "published": "2026-08-04T08:32:05Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03316", "title": "Any-OPD: Heterogeneous On-Policy Distillation for Flow-Matching Models via Representation-Space Bridging", "published": "2026-08-04T08:23:57Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "llm-architecture", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11249", "title": "Diffuse to Compress: Leveraging Diffusion LMs for Lossless Compression", "published": "2026-08-04T08:19:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "efficient-inference", "llm-architecture", "model-compression" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.03297", "title": "Distractor-Aware Truncation: Disentangling Context-Length Effects from Signal Loss in Long-Context LLM Benchmarks", "published": "2026-08-04T08:08:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03283", "title": "AgentPanel: Toward a New Paradigm for Human--AI Collaboration in Exploring Scientific Questions", "published": "2026-08-04T07:58:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03277", "title": "Noise-Aware Shrinkage for Differentially Private Zeroth-Order Fine-Tuning of Large Language Models", "published": "2026-08-04T07:52:23Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03272", "title": "Attacking and Defending Multi-Agent Collaborative Filtering Systems Through Connectivity", "published": "2026-08-04T07:49:58Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.03260", "title": "ED-DiT: Physics-Guided Diffusion Pretraining for Transferable Molecular Representations from Electron Density", "published": "2026-08-04T07:33:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03247", "title": "CIGTSurv: Clinical Information Guided Tri-modal Survival Prediction with Local Prototype Association and Global Feature Alignment", "published": "2026-08-04T07:19:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03239", "title": "Relational Priors as Convergence Pressure in LLM-Based Multi-Agent Systems", "published": "2026-08-04T07:11:28Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "multi-agent", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03222", "title": "Fail-Fast, Restart-Smart: Early Failure Prediction and Restart for SWE Agentic Tasks", "published": "2026-08-04T06:55:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03215", "title": "GROW: Group-Relative Advantage-Weighted On-Policy Reinforcement Learning of Autoregressive-Diffusion Text-to-Speech model", "published": "2026-08-04T06:50:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03214", "title": "The Agent Operating System (AOS): A Reference Operating Architecture for Distributed Agentic Systems", "published": "2026-08-04T06:50:49Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03204", "title": "Aligning Large Vision-Language Models at Test Time: A Trajectory-Guided Structured Sampling Approach", "published": "2026-08-04T06:47:08Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.03169", "title": "Test-time reasoning effort and unauthorized tool use in language-model agents: a prespecified equivalence study", "published": "2026-08-04T06:01:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03154", "title": "ANCHOR-RE: An Agentic Neuro-Symbolic Framework for Grounded Biomedical Relation Extraction", "published": "2026-08-04T05:36:19Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "post-training", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03138", "title": "Internalizing Academic Writing Workflows for Introduction Generation via Struct-Aware Policy Learning", "published": "2026-08-04T05:06:51Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03134", "title": "CLEAR: Causal Context-Based Agentic Reasoning for Vulnerability Detection", "published": "2026-08-04T05:03:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03105", "title": "HomoEnsNER: Does Language Alignment Outperform Architectural Complexity in Gujarati Named Entity Recognition?", "published": "2026-08-04T04:21:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03099", "title": "What Language Does and What the Evidence Supports: A Functional Role Taxonomy and Evidence Audit of Language Grounding in Embodied Agents", "published": "2026-08-04T04:17:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03095", "title": "VIVID: A Culturally Grounded Benchmark Exposing the Figurative Language Gap in Vietnamese NLP", "published": "2026-08-04T04:13:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03092", "title": "SMOPD: Multi-Reward Reinforcement Learning via Specialize-and-Merge Online Policy Distillation", "published": "2026-08-04T04:08:08Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation", "opd", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B08", "plan_reason": "OPD 与多教师/过程蒸馏 — completed" }, { "arxiv_id": "2608.03091", "title": "Position Bias Undermines Preference Consistency in Listwise LLM-Based Reranking", "published": "2026-08-04T04:04:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03089", "title": "Scalable Frequency- and Length-Aware Subdocument Deduplication for Large Language Model Pretraining", "published": "2026-08-04T04:02:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03083", "title": "GSTEP: Global Spatio-Temporal Density-Driven Visual Token Pruning for Efficient Video Large Language Models", "published": "2026-08-04T03:51:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03077", "title": "PAMT: Process-Aligned Reinforcement Learning for Multi-Domain Machine Translation", "published": "2026-08-04T03:42:54Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.03071", "title": "Getting the Parameters Right: A Difficulty-Graded Benchmark and Probe-Guided Training for LLM Tool Calls", "published": "2026-08-04T03:36:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03068", "title": "CVPO: Enhancing LLM Reinforcement Learning Reasoning via Value-Variance Adaptation and Dynamic Curriculum Learning", "published": "2026-08-04T03:30:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03063", "title": "SeqLLM: Augmenting LLMs with Behavioral-Sequence Modeling for High-Stakes Decisions at WeChat Pay", "published": "2026-08-04T03:23:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03062", "title": "TraceCAD: Trace-Guided Repair for Agentic CAD Generation", "published": "2026-08-04T03:23:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03048", "title": "PI-Mem: Pushing Long-Context Reasoning to 3.6M Tokens with Parallel-Iterative Memory", "published": "2026-08-04T02:59:16Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "llm-rl", "long-context", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.03044", "title": "Emulate or Estimate? The Divergent Strengths of Base and Post-Trained Language Models for Opinion Simulation", "published": "2026-08-04T02:51:38Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03041", "title": "PLAN: Parallel Liquid-Inspired Approximation Network for Efficient Representation Learning in Flexible Job Shop Scheduling", "published": "2026-08-04T02:43:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03036", "title": "LLM Serving in the Wild: An Empirical Study of Frameworks, Methods, and System Designs", "published": "2026-08-04T02:33:16Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-architecture", "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03031", "title": "CastFSR: A Fast--Slow--Reflect Agentic Reasoning Framework for Context-Aware Time Series Forecasting", "published": "2026-08-04T02:21:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03018", "title": "UrbanAgent: A Tool-Augmented Agent for Cross-System Urban Tasks", "published": "2026-08-04T02:04:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.03007", "title": "The Ground Is Shifting: A Reflection on the Foundations of Software Measurement", "published": "2026-08-04T01:41:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02993", "title": "Neurosymbolic Reasoning with Incremental Knowledge for Sample Efficient Hierarchical Reinforcement Learning", "published": "2026-08-04T01:17:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02986", "title": "Internalising the Identity Primitive: Cryptographic Individuality for an Autonomous Agent on a Public Blockchain", "published": "2026-08-04T00:47:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02985", "title": "Temporal Leakage in LLM Backtesting: Measurement, Validation, and Adjusted Scores", "published": "2026-08-04T00:45:54Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02961", "title": "Scaling an Autoregressive Transformer for Single-Cell Generation", "published": "2026-08-03T23:54:30Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02947", "title": "ATFlash: Per-RoPE-Wavelength Attention Windows for Compute/Memory-Efficient LLM Inference", "published": "2026-08-03T23:23:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02942", "title": "OPTD: On-Policy Transition Distillation with Consistency-Guided Adaptive Compression for Few-Step Diffusion Language Models", "published": "2026-08-03T23:09:43Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02919", "title": "FLARE: Few-shot Learning-based Adaptive Reflective Engine", "published": "2026-08-03T22:13:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02915", "title": "LACE: Large Language Model Aided Multi-Agent Framework for Agile RISC-V Instruction Extension", "published": "2026-08-03T22:07:06Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "multi-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02901", "title": "AnchorKV: Anchor-Residual KV Cache Compression", "published": "2026-08-03T21:38:30Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02878", "title": "VeriTrace: Human-Like Temporal Exploration Completes Agentic Action Space", "published": "2026-08-03T21:00:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18138", "title": "Language Models for Portuguese: A Systematic Mapping Study", "published": "2026-08-03T20:51:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02870", "title": "Maglev: Sliding Recurrent Memory", "published": "2026-08-03T20:40:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02867", "title": "BODHI: Do LLMs Branch Out and Discover Heterogeneous Inferences?", "published": "2026-08-03T20:37:54Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-rl", "reward-model", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.02831", "title": "Reinforcement Learning with Evolving Rubrics as Rewards for Audio Reasoning", "published": "2026-08-03T19:46:13Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.02826", "title": "Improved Quantum Algorithms for Reinforcement Learning Under a Generative Model", "published": "2026-08-03T19:38:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02807", "title": "Learning a Vector-Symbolic Model for Socio-Cultural Tasks", "published": "2026-08-03T19:03:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02602", "title": "AURORA-LM: Autoencoding Unified Representation for Continuous-Latent Diffusion Language Modeling", "published": "2026-08-03T17:59:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02712", "title": "Don't Regenerate, Debug: A Domain-Specific Agent for Repairing Near-Miss Hardware Operators", "published": "2026-08-03T17:59:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02595", "title": "onepot-Bench 0: towards lab-aware in silico chemistry benchmarks", "published": "2026-08-03T17:58:27Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02585", "title": "GradCuit: Credit-Assigned Gradient Flow Enables Robust and Interpretable Test-Time Latent Reasoning", "published": "2026-08-03T17:55:24Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02583", "title": "UEmbed: Unified Sparse and Dense Multimodal Embeddings", "published": "2026-08-03T17:54:11Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02582", "title": "ACEM: A Cost Estimation Model for Agentic Software Engineering", "published": "2026-08-03T17:54:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02569", "title": "AtumAI: A Principled Framework for Agentic Generation of Datacenter Control-Plane Policies", "published": "2026-08-03T17:45:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02560", "title": "Structured Memory for Edge Language Models: Persistent Context and Corpus Retrieval via O(1) SSM State Injection", "published": "2026-08-03T17:43:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "long-context", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.02553", "title": "A Taxonomy of Cognitive Capability Gaps in Generative and Agentic AI", "published": "2026-08-03T17:37:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02548", "title": "Between-User Collapse Under Popularity-Biased Feedback: A Centered-Covariance Theorem and Computable Phase Boundary", "published": "2026-08-03T17:33:25Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02508", "title": "RoMeRL: Balancing Feedback Coverage and the Memory-Reward Trap in Self-Evolving Agent Memory via Reduced-Order Utility States", "published": "2026-08-03T17:07:50Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agentic-rl", "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02499", "title": "SWE-Touch: Benchmarking Coding Agents When Users Touch the Code", "published": "2026-08-03T17:03:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02486", "title": "Cultural Awareness is Represented but Not Decoded: Tracing Mythological Knowledge across 18 Open-Source LLMs", "published": "2026-08-03T16:53:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02477", "title": "Unpaired Modality-Agnostic Generative Recommendation", "published": "2026-08-03T16:46:47Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02464", "title": "Real-Time Detection and Repair of LLM Agent Failures", "published": "2026-08-03T16:34:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16922", "title": "Towards welfare-oriented recommendations in activity-travel behavior", "published": "2026-08-03T16:34:35Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02422", "title": "Agentic Incident Response through Digital Twin-Enhanced Multiscale Planning", "published": "2026-08-03T16:03:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02407", "title": "Antares: Foundation Models for Agentic Vulnerability Localization", "published": "2026-08-03T15:49:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02391", "title": "Cooperative Coevolution for Resource-Constrained Agentic LLM Post-Training", "published": "2026-08-03T15:34:45Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agentic-rl", "llm-rl", "multi-agent", "post-training", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.02379", "title": "Chess on Ice: Curling Tactical Decision-Making via Backward Induction and Deep Reinforcement Learning", "published": "2026-08-03T15:23:42Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02376", "title": "Token-Native Storage: Read and Write in your Agent's Language", "published": "2026-08-03T15:20:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02372", "title": "PredAct-Bench: Benchmarking Tool-Augmented Dialogue under Controlled Tool Noise", "published": "2026-08-03T15:15:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02365", "title": "Faster-WAM: Do World Action Models Need Deep Action Modules?", "published": "2026-08-03T15:11:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02359", "title": "Fast and Accurate Quotation Attribution in Literary Texts", "published": "2026-08-03T15:08:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02358", "title": "ScrambleToolBench: Agents Search Exhaustively Even When Their Own Map Points to the Next Step", "published": "2026-08-03T15:07:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02353", "title": "Global Optimization and Inference-Time Region Grafting for Agentic Workflows", "published": "2026-08-03T15:04:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02352", "title": "Qwen-CUA: Native Computer Use for (almost) Everything", "published": "2026-08-03T15:04:20Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "reward-model", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02323", "title": "ECLAIR: A Causally-Grounded AI Framework for Scientific Discovery in Empirical Software Engineering", "published": "2026-08-03T14:47:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02703", "title": "ARCHead: Activation-Metric Residual Correction for Large Language Model Output Heads", "published": "2026-08-03T14:40:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02302", "title": "Trajectories That Segment Themselves: Agent-Declared Boundaries as a Training Unit", "published": "2026-08-03T14:27:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02287", "title": "SKT: Skill-Use Training at Scale via Verified Synthetic Data Generation", "published": "2026-08-03T14:18:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02276", "title": "Harness-R1: Learning to Edit Executable Runtime Harnesses from Agent Failure Trajectories", "published": "2026-08-03T14:12:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02254", "title": "Homebot: A Personal AI Agent for Conversational Home Assistance and Automation", "published": "2026-08-03T14:01:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02181", "title": "Start Classifying: Categorical Critics for LLM Reinforcement Learning", "published": "2026-08-03T13:05:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02698", "title": "Steganalysis of Adaptive Covert Collusion in Tool-Using Agent Populations: A Black-Box, Cross-Principal Approach", "published": "2026-08-03T13:00:25Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11248", "title": "EvoGraph-Mem: Failure-Aware Editable Graph Memory for Long-Term Language Agents", "published": "2026-08-03T12:35:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02139", "title": "Self-Improving Large Language Models via Progressive Experience Evolution", "published": "2026-08-03T12:27:32Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation", "post-training", "reward-model", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.02138", "title": "The Role of Disfluencies in Speech Translation", "published": "2026-08-03T12:26:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02113", "title": "MemArbiter: Decision-Time Memory Arbitration for Long-Horizon LLM Agents", "published": "2026-08-03T12:10:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02112", "title": "Do Static Embeddings Add Value to Hybrid Dutch Retrieval?", "published": "2026-08-03T12:10:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02110", "title": "IACM-RL: Intent-Aware Context Management and Reinforcement Learning for Complex Tool Invocation under Dynamic Intent Fluctuations", "published": "2026-08-03T12:09:50Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "llm-rl", "long-context", "model-compression", "on-policy-distillation", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.02101", "title": "Cross-Domain Hybrid OPD for Generalizable Search Agents", "published": "2026-08-03T12:01:18Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-architecture", "llm-rl", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02091", "title": "One QK Channel, Many Sources: Guarding Low-Precision Attention Collapse", "published": "2026-08-03T11:50:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02087", "title": "Instruction-Conditioned Exploration for Reinforcement Learning with Self-Distillation to an Unconditioned Policy", "published": "2026-08-03T11:47:52Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation", "post-training", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.02078", "title": "CAVE: Competence-Aware Visual Boundary Evidence Alignment for Video Temporal Grounding", "published": "2026-08-03T11:27:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02048", "title": "SmartGR: Hierarchy and Beam-Aware Knowledge Distillation for Generative Recommendation", "published": "2026-08-03T10:46:51Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02032", "title": "DART: Decoded Attention over Recurrent States for Efficient Long-Context Sequence Modeling", "published": "2026-08-03T10:30:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "long-context", "model-compression", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B07", "plan_reason": "LLM 架构、长上下文、KV cache 与评测基础设施 — completed" }, { "arxiv_id": "2608.02031", "title": "Learning-Based Collaborative MEC for LLM Inference with Soft-Deadline Awareness via Transformer-Enhanced PPO", "published": "2026-08-03T10:27:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "efficient-inference", "llm-architecture", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.02026", "title": "HPFA: Hypergraph-Based Paired Failure Attribution for LLM Reasoning", "published": "2026-08-03T10:23:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02693", "title": "PRWeaver: Evaluating LLM-Based Code Auditors against Long-Horizon Malicious Pull Requests", "published": "2026-08-03T10:05:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02001", "title": "VulnGym: Benchmarking Coding Agents for Repository-Level Vulnerability Detection", "published": "2026-08-03T10:03:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01995", "title": "Long-Horizon Autonomous Architecture Research with a Language-Model Agent: A Behavioural Case Study", "published": "2026-08-03T09:56:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01975", "title": "TELLER: Non-intrusive Cross-Layer Root-Cause Analysis for LLM Inference", "published": "2026-08-03T09:39:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01955", "title": "Agentic Self-Healing for Data and AI Pipelines: An Affordable Vendor-Agnostic Architecture using Open-Source Software", "published": "2026-08-03T09:26:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02691", "title": "Output-Aware Rotation for INT2 KV-Cache Quantization", "published": "2026-08-03T09:24:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01922", "title": "TRAM: Enhancing Multimodal Reasoning with Trajectory-Derived Auxiliary Memory", "published": "2026-08-03T08:57:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02689", "title": "Stuck on \"A\": Diagnosing and Repairing Interface Injury in Attention-to-KDA Linearization of a 0.6B Language Model", "published": "2026-08-03T08:54:18Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01913", "title": "Diagnosing Search Behavior and Failure Modes in Long-Horizon Search Agents", "published": "2026-08-03T08:46:48Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "long-context", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01904", "title": "CoEvoKG: Co-Evolving Knowledge Graphs with Self-Evolving Search Agents", "published": "2026-08-03T08:39:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01881", "title": "Hear, Invoke, and Understand: A Skill-Calling Multimodal Agent for Large Audio Language Models", "published": "2026-08-03T08:24:54Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01867", "title": "CRISP: Critical Step Perception for Training Efficient Deep Search Agents", "published": "2026-08-03T08:15:18Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "on-policy-distillation", "reward-model", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02687", "title": "PolicyGuard: Prompt-Configurable Semantic DLP for LLM Coding Agents", "published": "2026-08-03T08:10:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01851", "title": "Weights or Skills? A Survey of Robot-Learning Techniques: from Action-Predicting Weights to Robots that Write their Own Skills", "published": "2026-08-03T07:58:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01846", "title": "HyperAgent4POI: Dynamic Semantic Message Passing on Multi-Agent Hypergraphs for Missing-Modality Recommendation", "published": "2026-08-03T07:54:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01827", "title": "DeepVoyager-VL: Incentivizing Vision-in-the-Loop Search for Long-Horizon Multimodal Agents", "published": "2026-08-03T07:36:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01821", "title": "DAVET: Denoising-Aware Visual Evidence Trajectory Allocation for Diffusion Vision-Language Models", "published": "2026-08-03T07:29:05Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02685", "title": "BulkPR-Bench: Benchmarking Queue-Level Governance of Interacting Pull Requests", "published": "2026-08-03T07:24:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01804", "title": "LEAP: Lean Environment-Feedback via Adaptive Pruning for Code RL in GPU Kernel Generation", "published": "2026-08-03T07:12:52Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.01802", "title": "CoNav-UAV: Cooperative Dual-Altitude Aerial Navigation via Stackelberg Learning", "published": "2026-08-03T07:10:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01791", "title": "PICopilot: An LLM-based Agentic Framework for Assisting Photonic Integrated Circuit Design via Script Generation", "published": "2026-08-03T07:03:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01745", "title": "Heterogeneous Multi-Agent Reinforcement Learning for Radio Resource Management under Coupled Finite-Horizon Constraints", "published": "2026-08-03T06:11:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01744", "title": "RL-Lock: Reinforcement Learning for Generating Interlocking Assemblies", "published": "2026-08-03T06:11:17Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01743", "title": "Toward Plasticity-Preserving KL Regularization for Capability Retention in LLM Reinforcement Learning", "published": "2026-08-03T06:10:33Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01732", "title": "X-KGRank: A Knowledge Graph RAG Framework for Explainable Recommendations via Pattern Mining and LLM Re-Ranking", "published": "2026-08-03T05:56:40Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recommendation-ranking", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.01731", "title": "MODE: Mutual Optimality in Direct Effects of Reciprocal Recommendations in Matching Markets", "published": "2026-08-03T05:54:55Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01725", "title": "CENTILE: A Telemetry Foundation Model Evaluated by the Decisions It Drives", "published": "2026-08-03T05:45:19Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01719", "title": "MNC: Scope-Bound Semantic Declassification for Private LLM-Agent Communication", "published": "2026-08-03T05:36:49Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agent-planning", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.01717", "title": "Beyond On-Policy Exploration: Integrating External Policy Rollouts for Reinforcement Learning in Diffusion Language Models", "published": "2026-08-03T05:28:37Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B09", "plan_reason": "Rubric、外部 rollout 与多奖励 RL — completed" }, { "arxiv_id": "2608.01715", "title": "Coding Agents as Test-Suite Auditors: Finding What Official Suites Miss While Approaching What They Catch", "published": "2026-08-03T05:24:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01710", "title": "Beyond Single-Use Tokens: Durable Authorization State for Replay-Resistant LLM Agent Actions", "published": "2026-08-03T05:16:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01708", "title": "PGMem: Tightly Coupled Persona-Memory Graph for Lifelong Personalized Agents", "published": "2026-08-03T05:14:59Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01704", "title": "Floor, Ceiling, and the Fusion Gap: How Much of Crowd Reading Attention Can Machines Predict?", "published": "2026-08-03T05:09:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01684", "title": "GABench: A Comprehensive Benchmark for Evaluating LLM Agents on Graph Analysis Tasks", "published": "2026-08-03T04:18:31Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-planning", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.01678", "title": "Progressive Agent Skill Generation via Reinforcement Learning", "published": "2026-08-03T04:14:59Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.01676", "title": "Understanding Sparse Attention Selectivity in Long-Context Foundation Models via Counterfactual Evaluation", "published": "2026-08-03T04:12:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "long-context", "model-compression" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.01672", "title": "Learning What to Remember: Test-Time Training via Context Distillation", "published": "2026-08-03T04:06:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-architecture", "long-context", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B07", "plan_reason": "LLM 架构、长上下文、KV cache 与评测基础设施 — completed" }, { "arxiv_id": "2608.01667", "title": "TCPO: Turn-Level Credit Policy Optimization", "published": "2026-08-03T04:01:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01664", "title": "FAU at ImageCLEF 2026 Task on Multimodal Reasoning Robust Candidate Scoring and Concise Multilingual Visual Answering", "published": "2026-08-03T03:54:22Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01662", "title": "LongCat Sparse Attention: Taming the Lightning via Streaming-aware Hierarchical Cross-Layer Indexing", "published": "2026-08-03T03:51:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01651", "title": "Bole: Efficient Tree Speculation for Hybrid-Attention Language Models", "published": "2026-08-03T03:43:14Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "efficient-inference", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01648", "title": "Evaluating Forecasting Techniques for Hardware Errors on a Large-scale HPC System", "published": "2026-08-03T03:38:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01645", "title": "GISAgentBench: A Practitioner-Sourced Benchmark for Evaluating LLM Agents on GIS Tasks", "published": "2026-08-03T03:31:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01633", "title": "GraphIR: Architecture-Level Search States for LLM-Guided Neural Architecture Evolution", "published": "2026-08-03T03:06:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01631", "title": "Does Accuracy Equal Evidence? Reasoning Faithfulness under KV Cache Compression", "published": "2026-08-03T03:03:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01630", "title": "RING: Retrieval-Internalized Generation for Continual Large-Scale Knowledge Injection", "published": "2026-08-03T03:00:43Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01604", "title": "Post-Training on Office Work Improves Software Engineering: A Behavioral Account of Cross-Domain Transfer", "published": "2026-08-03T02:22:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02683", "title": "$S^3$: Improving Agent Safety through Multi-Stage Defense", "published": "2026-08-03T02:06:06Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agent-planning", "multi-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.01575", "title": "Measuring in-context algorithmic reasoning in language models against an exact Bayes-optimal standard", "published": "2026-08-03T01:21:44Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02680", "title": "TraceCompiler: Skill-Guided Mining and Compilation of LLM Agent Traces into Mostly Deterministic Workflows", "published": "2026-08-03T01:05:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01559", "title": "Does the Competitive Component of Adversarial Self-Play Improve Legal Reasoning? A Controlled Negative Result", "published": "2026-08-03T00:31:34Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01558", "title": "Securing Agentic AI: From Per-Action Checks to Trajectory Assurance", "published": "2026-08-03T00:28:49Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.01556", "title": "Rethinking Personalized Reward Modeling for LLMs under Preference Heterogeneity via Group-Debiased Federated Learning", "published": "2026-08-03T00:25:52Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01536", "title": "Celty: SpMspV GPU Kernel and SIMT Co-Design for Efficient Dual-Sparse LLM Inference", "published": "2026-08-02T23:10:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "efficient-inference", "llm-architecture", "model-compression" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.01522", "title": "Question Begets Question: Self-Evolving Curriculum for Reinforcement Fine-Tuning on Competition Mathematics", "published": "2026-08-02T22:19:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01507", "title": "Deep Agentic Search for Repository-Level Code Question Answering: An Empirical Study", "published": "2026-08-02T21:36:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.04032", "title": "EDATracer: An Agentic Framework for Large-Scale EDA Artifact Analysis", "published": "2026-08-02T20:06:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16919", "title": "CARA: Cognitive Adaptive Recommendation Agent", "published": "2026-08-02T19:57:59Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01460", "title": "Conformalized Large Language Models under Configuration Shift", "published": "2026-08-02T19:34:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01458", "title": "PALMs: Using Multi Construct-Grounded Rationales for Modeling Population Preferences in LLMs", "published": "2026-08-02T19:31:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01425", "title": "Training Small LLMs as Spatial Multi-Agent Policies", "published": "2026-08-02T18:14:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01418", "title": "Reusing Rollouts under Policy Lag: Prefix-Normalized Policy Optimization for LLM Reinforcement Learning", "published": "2026-08-02T18:02:03Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01395", "title": "Language Equality has a Price: A Systematic Investigation of Multi-turn LLM Performance for EU-24+", "published": "2026-08-02T17:20:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01366", "title": "Asking Questions the Right Way: A Multi-Agent Conversational System for Prompt Formulation in Complex Task Resolution", "published": "2026-08-02T16:33:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01359", "title": "EviSD: Evidence-Conditioned Self-Distillation for Search-Augmented Agents", "published": "2026-08-02T16:22:27Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.01358", "title": "HopRefusalBench: Diagnosing Refusal Failures in Search-Augmented Agents for Multi-Hop Reasoning", "published": "2026-08-02T16:20:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01352", "title": "Spatiotemporal Proximal Causal Inference under Hidden Confounding and Interference", "published": "2026-08-02T16:13:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01347", "title": "Prompt-Induced Waste in Coding Agents: Reasoning, Effort, Harness Design, and End-to-End Cost", "published": "2026-08-02T16:10:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01328", "title": "LongChart VQA: A Comprehensive Benchmark for MLLMs with Complex Multi-Chart Reasoning", "published": "2026-08-02T15:48:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01321", "title": "BiCAA: Bidirectional Credit Assignment for Search-Augmented Agent", "published": "2026-08-02T15:41:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01315", "title": "Collaborative Memory Augmentation for Generative Recommendation", "published": "2026-08-02T15:34:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01311", "title": "RH-RAG: Trustworthy Long-Form Generation for Privacy-Constrained Settings", "published": "2026-08-02T15:27:00Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning", "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01298", "title": "UDT: Reconciling U-Nets and Diffusion Transformers with Data-Adaptive Token Reduction", "published": "2026-08-02T15:07:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01285", "title": "Stop When Memory Suffices: Evidence-Conditioned Progressive Execution for LLM Agents", "published": "2026-08-02T14:49:24Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01284", "title": "Training nGPT", "published": "2026-08-02T14:47:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01283", "title": "Riemannian Attention Mechanisms for Transformers: A Theoretical Framework and Architecture Design", "published": "2026-08-02T14:47:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01263", "title": "Distill What the Student Can See: Fisher-Projected On-Policy Distillation for Vision-Language Models", "published": "2026-08-02T14:16:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01260", "title": "Auditing Semantic Gains in Sequential Recommendation: A Lightweight Recovery Test", "published": "2026-08-02T14:15:50Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01247", "title": "RestoreKV: Recovering Full-Cache Behavior Under Aggressive Query-Agnostic KV Cache Eviction", "published": "2026-08-02T13:58:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01234", "title": "Learning What to Remember and What to Internalize in LLM Self-Evolution via Adaptive Memory-Parameter Coordination", "published": "2026-08-02T13:39:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01205", "title": "ReBRAC-v2: The Return of the King", "published": "2026-08-02T12:42:20Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01202", "title": "Fruit-HSNet: A Machine Learning Approach for Hyperspectral Image-Based Fruit Ripeness Prediction", "published": "2026-08-02T12:37:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01194", "title": "Hybrid Quantum Neural Networks: Theory, Implementations, and Applications", "published": "2026-08-02T12:21:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01193", "title": "Humans Are More Diverse: Frontier LLMs Show Extreme Policies in Idealised AI Development Races", "published": "2026-08-02T12:18:00Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01189", "title": "MADE: Belief-Driven Dual-Agent Coordination for Autonomous Model Deployment", "published": "2026-08-02T12:14:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01185", "title": "3DZip: Spatial-Aware Feature Diversity-Guided Token Compression for 3D Question Answering", "published": "2026-08-02T12:11:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01184", "title": "SAFE-Merge: Data-Free Continual Model Merging with General Knowledge Preservation", "published": "2026-08-02T12:09:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01147", "title": "UniHEAR: Unified Heterogeneous-Source Attentive Retrieval for Knowledge-Based Visual Question Answering", "published": "2026-08-02T10:57:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11244", "title": "BEST-KAG: Enhancing Question Answering of Building Engineering Standards with Multimodal Knowledge Graph Modeling and Large Language Model", "published": "2026-08-02T10:27:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01133", "title": "Policy Optimality Measurement for Multi-Vehicle Decision-Making: From Extrinsic Indicators to Intrinsic Quality", "published": "2026-08-02T10:21:58Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "agentic-rl", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01128", "title": "MA-HEAD-Net: Adaptive Rule-Guided Multi-Agent DRL for AoI Minimization in UAV-Assisted Emergency Networks", "published": "2026-08-02T10:05:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01078", "title": "Attend to Your Own Thoughts: Breaking the Barrier for Post-Training Quantization of Reasoning LLMs through the Lens of 1.58-Bit Quantization", "published": "2026-08-02T08:22:43Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01074", "title": "Logit-Origin Centering for Singleton Test-Time Adaptation", "published": "2026-08-02T08:18:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01056", "title": "Control Under Compression: Reliability Frontiers for Tool-Using Agents", "published": "2026-08-02T07:43:44Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01050", "title": "Don't Offer What Can't Be Done: Deterministic Executability Gating for LLM Skill Selection at Scale", "published": "2026-08-02T07:32:49Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01042", "title": "What Could the Agent See at 19:05? Generating Temporal Enterprise Scenarios from Real Research and Replaying Them to Evaluate Agents", "published": "2026-08-02T06:56:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01034", "title": "Opt.Gear Technical Report", "published": "2026-08-02T06:43:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01014", "title": "Cloud-ScPO: Hidden-State Geometry for Semi-Supervised Preference Optimization in LLM Reasoning", "published": "2026-08-02T05:33:41Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01012", "title": "MedUPS: Towards Diagnostic Assistance in Uncommon Medical Cases with Large Language Models", "published": "2026-08-02T05:27:21Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01004", "title": "Who Belongs in the Eval Set? A Capability-Taxonomy-Driven Pipeline for Curating Regression Eval Sets in Agent-Extensibility Platforms", "published": "2026-08-02T05:19:25Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.01001", "title": "From AI Technical Debt to Agentic Technical Debt: A Systematic Mapping of Root Causes and Manifestations in Agentic AI Systems", "published": "2026-08-02T05:09:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00974", "title": "Search-GRT: Guided Retrieval Training of Search Agents to Optimize for Complex Question Answering", "published": "2026-08-02T03:55:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00969", "title": "PROGRESS: Coverage-guided RL to Train Search-augmented LLM Agent", "published": "2026-08-02T03:48:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00967", "title": "TrajWiki: Source-Grounded Memory Trajectories for Long-Horizon Dialogue Agents", "published": "2026-08-02T03:36:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00966", "title": "AgenTag: Attribution of AI Coding Agents from Behavioral Fingerprints", "published": "2026-08-02T03:35:25Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00947", "title": "Claim Plane: Reliability Gains and the Limits of Selective Concurrency for Parallel Coding Agents: A 30-Pair, Three-Seed Confirmatory Study of Deterministic Pre-Write Admission", "published": "2026-08-02T02:56:49Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00938", "title": "GRACE: Generative Recommender Acceleration Engine for Real-Time Ads Retrieval", "published": "2026-08-02T02:30:38Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.00937", "title": "Neuro-Symbolic Participation Governance for Verifiable AI Agents in Open Digital Twin Ecosystems", "published": "2026-08-02T02:30:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00932", "title": "Gaokerena: A Small Persian Medical Language Model Family", "published": "2026-08-02T02:17:10Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture", "llm-rl", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00924", "title": "RefactorAssist: Agentic Refinement for Reliable Code Refactoring", "published": "2026-08-02T01:32:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00916", "title": "Tevatron Meets Megatron: Expert-Parallel LLM Reranker Training on an Academic Budget", "published": "2026-08-02T00:55:47Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00909", "title": "FinHardBench: Can LLMs Generate Latency-Aware Hardware for Financial Computing?", "published": "2026-08-02T00:30:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00902", "title": "Practical Online KV Cache Compaction for LLM Agents: An Empirical Study", "published": "2026-08-02T00:08:44Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00852", "title": "HyperODE: Zero-Shot Surrogate for Simulation and Inference of Dynamical Systems", "published": "2026-08-01T20:16:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00837", "title": "Pruned BPE: Post-training Visibility Pruning and Token Reallocation for Byte Pair Encoding", "published": "2026-08-01T19:34:44Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00835", "title": "Deep Learning CNN and Recurrence Analysis for Alpha Gamma EEG Biomarkers in Fragile X Syndrome", "published": "2026-08-01T19:27:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02663", "title": "CT-HEG: A Bidirectional, Timestamp-Attributed Event Graph for ICU In-Hospital Mortality Prediction - An Architectural Ablation Study", "published": "2026-08-01T19:10:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00816", "title": "Exponential Reward Weighting for Fine-Tuning Generative Recommenders under Sparse and Noisy Feedback", "published": "2026-08-01T18:48:39Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "industrial-ranking", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.00808", "title": "Turning Interaction History into Execution State: A Runtime Layer for Long-Horizon Coding Agents", "published": "2026-08-01T18:25:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00796", "title": "An Uncertainty-Driven Hybrid Deep Learning Approach for Broad-Coverage RF Modulation Recognition", "published": "2026-08-01T17:57:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00782", "title": "Distill Where You Fail: Recovering Learning Signals of Negative RL-Groups from Adaptive Teacher Guidance", "published": "2026-08-01T17:29:14Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B08", "plan_reason": "OPD 与多教师/过程蒸馏 — completed" }, { "arxiv_id": "2608.04030", "title": "NuclearDiffusion: Text-to-Image Foundation Models for Learning Nuclear Energy Concepts", "published": "2026-08-01T17:14:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00765", "title": "RAGOCR: Optical Compression of Retrieval-Augmented Text via Visual Representation", "published": "2026-08-01T16:59:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00747", "title": "When Prompts Control Robots: Prompt Injection Attacks in Multi-Agent Robotic Systems", "published": "2026-08-01T16:31:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07550", "title": "Auditing Medical Vision-Language Models on Chest Radiographs: Estimating Reference Agreement Across Institutions", "published": "2026-08-01T16:18:59Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00722", "title": "Experience-Calibrated Contrastive Decoding for Mitigating Hallucinations in LM-Based Text-to-Speech", "published": "2026-08-01T15:41:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00718", "title": "Adversarial Attacks in Multi-Agent LLM Pipelines: Unveiling Structural Vulnerabilities in Agentic AI Architectures", "published": "2026-08-01T15:35:58Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agent-planning", "multi-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.00713", "title": "Observatorio Lazaro: A self-populating database of anglicism usage in the Spanish press", "published": "2026-08-01T15:21:59Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00712", "title": "Exploiting Intrinsic Duality for Multi-Hop Question Generation", "published": "2026-08-01T15:20:16Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-architecture", "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00711", "title": "Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations", "published": "2026-08-01T15:19:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00693", "title": "AttnLink: Turning Attention into Schema Links for Text-to-SQL", "published": "2026-08-01T14:44:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00692", "title": "Vul4Py: Benchmarking Automated Vulnerability Repair in Python with Paired Exploit and Functional Oracles", "published": "2026-08-01T14:40:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00679", "title": "HetGPS: Scalable Graph Multi-Agent Reinforcement Learning with Physics-Anchored Adaptive Safety for EV Charging", "published": "2026-08-01T13:59:49Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00669", "title": "GARDRec: Decision-Level Graph Grounding for Large Language Model Recommendation", "published": "2026-08-01T13:44:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00640", "title": "TreeProbe : A Tibetan Medicine Benchmark for Cultural Bias in LLMs", "published": "2026-08-01T12:53:24Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00584", "title": "Element-Aware Group Learning for E-Commerce Image Generation", "published": "2026-08-01T10:45:03Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00583", "title": "A False Average: Chain-of-Thought Monitors Collapse Where They Are the Only Defense", "published": "2026-08-01T10:42:41Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00582", "title": "Writing-System-Level Tokenizer Adaptation for Byte-Level BPE", "published": "2026-08-01T10:38:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00577", "title": "HetRoute Heterogeneous and Cost-aware Collaborative Routing Framework for Distributed Edge MoE Inference", "published": "2026-08-01T10:30:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00573", "title": "TrimMoE A communication aware and adaptive depth framework for distributed edge inference", "published": "2026-08-01T10:19:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00558", "title": "AiFlow: Token-Native Reactive Orchestration with Bounded Backpressure for Streaming LLM Applications", "published": "2026-08-01T09:33:29Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "multi-agent", "software-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.00542", "title": "Agentic Graph Token Reasoning", "published": "2026-08-01T08:56:26Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "process-reward", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00533", "title": "Native Multilingual Chain-of-Thought Reasoning in Low-Resource Southeast Asian Languages", "published": "2026-08-01T08:44:32Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00528", "title": "S$^4$R: Selective Sampling, Subspaces, and Sparse Reconstruction for Compressed Long-Context KV Caching", "published": "2026-08-01T08:41:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00507", "title": "The Learning Objective Governs Perceptual Narrowing: A Cross-Lingual, Layer-Wise, Ten-Seed Study of Self-Supervised Speech Encoders", "published": "2026-08-01T08:02:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00494", "title": "TaPR: Test-Aware Policy Refinement for Feedback-Conditioned Code Generation", "published": "2026-08-01T07:35:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00473", "title": "CrossProjection: Geometric Grounding Beyond Viewpoint Change in Architectural Drawings", "published": "2026-08-01T06:51:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00450", "title": "Revibing Code from Papers: Reimplementing HCI Artifacts", "published": "2026-08-01T05:18:25Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00432", "title": "Deep Research Pretraining via Predictive Navigation", "published": "2026-08-01T04:17:14Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data", "tool-agent", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00419", "title": "Unleashing the Potential of Large Language Models: A Blueprint for Real-Time, Enterprise-Ready Deployments", "published": "2026-08-01T03:40:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00335", "title": "RMSWeb: Reflection, Failure-Mode Mining, and Salvage-DS for Web Agent Reinforcement Learning", "published": "2026-07-31T22:57:16Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "post-training", "reward-model", "web-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.00326", "title": "Learning to Coordinate Symbolic Tools: LLM Agents for Verified Sum-of-Squares Certificates", "published": "2026-07-31T22:26:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00311", "title": "SeDeM: Selective Decompression of Hidden-State Memories for Long-Context Question Answering", "published": "2026-07-31T21:44:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02650", "title": "HyperAgent: Planning and Acting over Tool-Schema Hypergraphs for Tool-Use LLM Agents", "published": "2026-07-31T21:42:35Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-planning", "software-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B11", "plan_reason": "Agent 记忆、工具规划与自进化系统 — completed" }, { "arxiv_id": "2608.00303", "title": "CrystalMem: Elastic Memory for Self-Evolving LLM Agents via Knowledge Crystallization", "published": "2026-07-31T21:35:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00301", "title": "Abstention as an Action Can Kill Both the Reward Gradient and the KL Anchor: Collapse Law and Repair for Error-Penalized Reinforcement Learning", "published": "2026-07-31T21:31:07Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00267", "title": "LoopsBench: From Harness Engineering to Loop Engineering in Coding Agent Evaluation", "published": "2026-07-31T20:18:25Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00243", "title": "More Debate, Same Evidence: Structural Limits of Homogeneous Multi-Agent Groundedness", "published": "2026-07-31T19:40:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.05188", "title": "Position: It's Time to Optimize LLMs for Self-Consistency", "published": "2026-07-31T19:16:53Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00220", "title": "Verifier-Induced Support Reshaping in On-Policy Optimization", "published": "2026-07-31T19:05:43Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.00218", "title": "A Few Neurons Reveal When LLMs Misuse Tools: Sparse Detection and Selective Steering for Reliable Tool Use", "published": "2026-07-31T19:03:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00215", "title": "Personalizing Large Language Model Agents with Small Policy Models", "published": "2026-07-31T18:56:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13598", "title": "Measuring Cross-Task Behavioral Consistency in Language Model Agents", "published": "2026-07-31T18:37:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00181", "title": "Cross-Benchmark Generalization in Long-Horizon Agents", "published": "2026-07-31T18:05:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00175", "title": "Inference-Time Policy Alignment for Fair Reinforcement Learning", "published": "2026-07-31T18:00:50Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "preference-optimization", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.00155", "title": "AgentStream: How Well Do Self-Evolving LLM Agents Perform Under Streaming Tasks?", "published": "2026-07-31T17:50:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29658", "title": "Reusing Past Repairs Through Hierarchical Trajectory Abstraction for Coding Agents", "published": "2026-07-31T17:38:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29637", "title": "CodeShrink: Adaptive Visual Compression for Efficient Multimodal Code Understanding", "published": "2026-07-31T17:15:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29617", "title": "When Does On-Policy Interaction Help? Representational Tradeoffs in Value-Based Imitation Learning", "published": "2026-07-31T16:52:47Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29613", "title": "WCM: A World Critic Model for Vision-Language-Action Reinforcement Learning", "published": "2026-07-31T16:48:45Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29610", "title": "Educating the Agentic Engineer: Curricula, Collaboration, and Continuous Learning in the AI Era", "published": "2026-07-31T16:45:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02645", "title": "Verified Tool Calls Improve LLM Agent Reliability Under Non-Atomic Failures", "published": "2026-07-31T16:16:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00146", "title": "DiffusionGemma Technical Report", "published": "2026-07-31T16:11:46Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29577", "title": "DungeonBench: A Benchmark for Rules-Rich Tactical Reasoning in Dungeons & Dragons Combat", "published": "2026-07-31T16:03:38Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29559", "title": "LEMUR: Learning to Align with Multi-Objective Reinforcement Learning from Preference Feedback", "published": "2026-07-31T15:50:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11241", "title": "RecSys Factory: Bounding LLM Agent Autonomy to Decision Points in the Industrial Recommender Lifecycle", "published": "2026-07-31T15:42:45Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-tencent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.29516", "title": "From Code Review to Code Critique: Intent, Drift, and Spotlight for AI-Generated Diffs at Scale", "published": "2026-07-31T15:17:25Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29494", "title": "Adaptive FastOPD: Progress-Aware Rollout Horizon Expansion for Efficient On-Policy Distillation", "published": "2026-07-31T15:02:53Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29484", "title": "Evidence-Type Competition: When Can Interventional Data Teach Language Models Causal Direction?", "published": "2026-07-31T14:49:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02643", "title": "CUADebug: Diagnosing and Repairing Computer-Use Agent Failures", "published": "2026-07-31T13:59:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29433", "title": "Know It, Act on It: Investigating Memory Utilization in LLM Personalization", "published": "2026-07-31T13:57:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.06397", "title": "Agentic Planning for Symbolic Execution", "published": "2026-07-31T13:47:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29422", "title": "AgenticRepair: Multi-Faceted Program Context Engineering for Agentic Vulnerability Repair", "published": "2026-07-31T13:42:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29419", "title": "Explore Beyond the Boundary Using Entropic Information", "published": "2026-07-31T13:41:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00130", "title": "A Fortran General-Purpose Transpiler: Proof of Concept", "published": "2026-07-31T13:38:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02641", "title": "IR2Solve: Structured Intermediate Representations for Cost-Efficient Optimization Autoformulation", "published": "2026-07-31T13:36:26Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29405", "title": "Beyond Component Testing: Validating Agentic AI Systems", "published": "2026-07-31T13:24:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29397", "title": "Studying quantization trade-offs for efficient inference deployment in machine translation", "published": "2026-07-31T13:15:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29378", "title": "PTP: Previous-Token Prediction based LLM Inversion for Near-Exact Prompt Reconstruction", "published": "2026-07-31T13:01:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29363", "title": "Stable Autoregressive Speech Generation with Low-Frame-Rate High-Dimensional Continuous Tokens", "published": "2026-07-31T12:51:24Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14658", "title": "pico-type: A 1.5M-Parameter Byte-Level Multi-Head Content Classifier", "published": "2026-07-31T12:20:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29320", "title": "MAGA: Multi-Platform Self-Fusion of GUI Agents via Structured Action Distillation", "published": "2026-07-31T11:51:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29287", "title": "Translation with Thought: Difficulty-Adaptive Reasoning via Reinforcement Learning for Multi-Domain Machine Translation", "published": "2026-07-31T10:59:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02638", "title": "Studying, Identifying, and Fixing Hidden Technical Debt in AI-Intensive Cyber-Physical Systems", "published": "2026-07-31T10:44:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29256", "title": "UniPolymer: A Unified Framework for Property Prediction, Structure Recommendation, and Evaluation in Polyimide Design", "published": "2026-07-31T10:29:54Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29254", "title": "Tool Specifications Matter: Uncovering and Mitigating Safety Risks in AI Agents", "published": "2026-07-31T10:25:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29250", "title": "Data Turnstile: A Scalable Open Framework for Function-Calling Data Generation", "published": "2026-07-31T10:21:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29241", "title": "RecHarness: A Bandit-Routed Agentic Harness for Self-Evolving Recommender Systems", "published": "2026-07-31T10:15:38Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "production-evidence", "recsys-general" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B02", "plan_reason": "7 月工业生成推荐、Agent harness 与搜索 — completed" }, { "arxiv_id": "2607.29213", "title": "GALA: Generative Aligned Learning for Adaptive Multimodal Representation in the Taobao Shangou Recommender System", "published": "2026-07-31T09:32:01Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "generative-recommendation", "industrial-ranking", "priority-org-taobao", "production-evidence", "recsys-general", "search-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B02", "plan_reason": "7 月工业生成推荐、Agent harness 与搜索 — completed" }, { "arxiv_id": "2607.29211", "title": "Knowing When to Quit: Diagnosing and Training LLMs to Abort Futile Reasoning", "published": "2026-07-31T09:30:33Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29196", "title": "Hy-MultiTurn: A Six-Dimensional Benchmark for Deep Multi-Turn Dialogue Understanding", "published": "2026-07-31T09:15:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29185", "title": "Learning Latent Reasoning Traces for Scalar Reward Models End-to-End", "published": "2026-07-31T09:05:58Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "preference-optimization", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.29175", "title": "Execution-First Synthetic Tool-Use Trace Generation for LLM Agents", "published": "2026-07-31T08:52:26Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29167", "title": "Memory Provenance Laundering in LLM Agents: A Non-Amplification Firewall for Persistent Memory", "published": "2026-07-31T08:46:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29082", "title": "Can Zero-Shot LLMs Predict Child Malnutrition? A Fairness and Temporal Robustness Study", "published": "2026-07-31T07:06:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29071", "title": "Federated Foundation Models Fine-Tuning with Heterogeneous Compressed Clients", "published": "2026-07-31T06:45:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07545", "title": "DarwinX: Evolving Agent Harnesses Through Natural Selection", "published": "2026-07-31T05:34:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29032", "title": "TransMem: Transforming Hidden States into Memory for Large Language Models", "published": "2026-07-31T05:11:04Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-memory", "efficient-inference", "long-context", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B07", "plan_reason": "LLM 架构、长上下文、KV cache 与评测基础设施 — completed" }, { "arxiv_id": "2607.29010", "title": "EvoReason: Self-Evolving Reasoning Primitive-Guided On-Policy Distillation for Latent Reasoning in Generative Recommendation", "published": "2026-07-31T04:24:16Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02636", "title": "Rethinking Self-Evolving Agent Skills: Feedback Dynamics over Multiple Rounds", "published": "2026-07-31T04:02:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.29002", "title": "MMShopBench: A Real-Log Benchmark for Multimodal, Multi-Turn Shopping Agents", "published": "2026-07-31T04:00:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28997", "title": "Think2Go: Generative Next POI Recommendation with LLM Reasoning", "published": "2026-07-31T03:52:20Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28990", "title": "Scaling Scientific Discovery Environments for Turn-Level Agentic RL", "published": "2026-07-31T03:34:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28986", "title": "Adjudicated Captioning: Multi-Agent Alignment Scoring and Consensus-Distilled Beam Arbitration for Strict Zero-Shot Image Captioning", "published": "2026-07-31T03:27:57Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28982", "title": "PARALLEL: A Prefrontal-Aligned Reinforcement inspired Approach for Language-Model Learning under Explicit Limits", "published": "2026-07-31T03:20:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28979", "title": "Mixture-of-Translators: Translating KV Caches Across Heterogeneous Large Language Models", "published": "2026-07-31T03:07:31Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "long-context", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28971", "title": "Don't Contrast the Impossible: Region-Constrained Batching for Contrastive User Modeling on a Local Community Platform", "published": "2026-07-31T02:47:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "production-evidence", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.11232", "title": "Backtrader-Bench: Benchmarking LLM Agents on Algorithmic Trading with Self-Generated MCQs", "published": "2026-07-31T02:37:57Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28966", "title": "BLADE: Boundary-Expanded and Layer-Adaptive Dynamic Exit for Efficient LLM Reasoning", "published": "2026-07-31T02:36:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28942", "title": "NeSyFS: A Neuro-symbolic Fast-Slow Thinking Framework for LLM Agent under Partial Observability", "published": "2026-07-31T01:53:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28940", "title": "TransX: Scaling Transformer-based Recommendation via Behavioral and Serving Stream Crossings", "published": "2026-07-31T01:42:57Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.28928", "title": "Automated Testing and Repair for Verified Compilers Generated by a Coding Agent", "published": "2026-07-31T01:04:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28916", "title": "Gated Q-learning: Add Off-Policy Bias to Taste", "published": "2026-07-31T00:32:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28895", "title": "LLM-Based Generative Retrieval for Snapchat Content Recommendation", "published": "2026-07-30T23:27:14Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.28889", "title": "Human-LLM Collaborative Inductive Coding for Conceptualizing K-12 Educator AI Use", "published": "2026-07-30T23:16:20Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28887", "title": "To Add Is Machine, To Delete Is Human: Measuring and Mitigating Deletion Avoidance in LLM Code Editing", "published": "2026-07-30T23:07:51Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28877", "title": "Open-Source LLM-Driven Formal Verification: A Multi-Agent Pipeline for RTL Repair", "published": "2026-07-30T22:43:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28871", "title": "Validation Evidence in LLM Repair Agents: How Much of What Passes Actually Tests the Bug?", "published": "2026-07-30T22:32:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28862", "title": "TextCloak: Thwarting Unauthorized LLM Exploitation via RL-Driven Unlearnable Text", "published": "2026-07-30T22:01:36Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28841", "title": "CyberNeuro: A Privacy-Preserving Agentic Workbench for Cohort-Scale Neuroimage and Clinical Data Analysis", "published": "2026-07-30T21:10:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "software-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28840", "title": "Benchmarks Are Not Validation: A System-Level View of Financial LLM Applications", "published": "2026-07-30T21:07:27Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "multi-agent", "software-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.00101", "title": "Agentic Coding in the Wild: Characterizing GitHub Copilot Traces at Production Scale", "published": "2026-07-30T20:51:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28829", "title": "When Unlearning Fails: Reliable Data Deletion under Post-Training in Agent Networks", "published": "2026-07-30T20:41:53Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28826", "title": "Distilling Knowledge from Large Language Models into Lightweight Reinforcement Learning Agents for Autonomous Cyber Operations", "published": "2026-07-30T20:28:02Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "on-policy-distillation", "opd", "pretraining-data", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.28818", "title": "Best Friends, Not Forever: Evaluating Long-Horizon Persona Collapse and Behavioral Drift in AI Companions", "published": "2026-07-30T20:18:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28814", "title": "Rolling With Resistance: Preference-Optimized LLM Counselors Can Trade Goal Persistence for Relational Attunement in Motivational Interviewing", "published": "2026-07-30T20:16:03Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28777", "title": "Self-Supervised Skill Optimization", "published": "2026-07-30T19:04:34Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28628", "title": "Learning to Trace Seiberg Dualities", "published": "2026-07-30T17:59:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28589", "title": "MixFrag: Fragility-Guided Mixed-Precision Post-Training Quantization for Vision Transformers", "published": "2026-07-30T17:43:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28568", "title": "Frontis-MA1: Training an AI4AI Model towards Recursive Self-Improvement in Machine Learning Engineering", "published": "2026-07-30T17:34:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28527", "title": "MANTA: Multi-Agent Network Topology Adaptation for Self-Evolving Multi-Agent Systems", "published": "2026-07-30T17:01:27Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-planning", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B11", "plan_reason": "Agent 记忆、工具规划与自进化系统 — completed" }, { "arxiv_id": "2607.28496", "title": "Beyond Sentiment: Structured Information Extraction from Financial News", "published": "2026-07-30T16:41:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28476", "title": "Improving Mental Health Screening and Early Risk Detection in Spanish", "published": "2026-07-30T16:28:28Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "long-context", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28460", "title": "Cybersecurity Detection Classification with Reasoning-enabled Language Models", "published": "2026-07-30T16:22:13Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28457", "title": "SVR: Self-Verifying Refinement via Joint Verdict-Confidence Reinforcement Learning for Adaptive Test-Time Compute", "published": "2026-07-30T16:20:58Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "reward-model", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.28449", "title": "Lightning OPD 2.0: Mitigating Style Bias in Cross-Teacher On-Policy Distillation for Large Reasoning Models", "published": "2026-07-30T16:17:15Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28707", "title": "Demystifying Entropy-based Selection for Chain-of-Thought Compression in Large Reasoning Models", "published": "2026-07-30T15:59:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28390", "title": "Hierarchical Multilevel Monte Carlo for Order-Optimal Neural Actor-Critic in Average-Reward CMDPs", "published": "2026-07-30T15:46:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28304", "title": "Semi-Supervised Learning for Molecular Graphs via Ensemble Consensus", "published": "2026-07-30T14:43:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28301", "title": "HARGO: Heterogeneity-Aware Reward-Guided Optimization for RL Post-Training of LLMs on HPC Tasks", "published": "2026-07-30T14:41:05Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.28272", "title": "MemHarness: Memory Is Reconstructed, Not Replayed", "published": "2026-07-30T14:25:49Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28263", "title": "Understanding Is Done Early: A Depth Division of Labor in Large Language Models and Its Use for Unbounded-Context Memory", "published": "2026-07-30T14:19:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28259", "title": "TopoFormer: Topology Meets Attention for Graph Learning", "published": "2026-07-30T14:18:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28212", "title": "Causal Discovery with Inverted Self-attention for Multivariate Time Series", "published": "2026-07-30T13:49:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28200", "title": "Vibe-FDTR: An agent-oriented framework for reproducible frequency-domain thermoreflectance data analysis", "published": "2026-07-30T13:38:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28196", "title": "Fidelity Is Not Safety: Gently-Compressed LLMs Pass Every Data-Free Quality Guard Yet Invent Procedure Steps in Agentic Execution", "published": "2026-07-30T13:33:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28182", "title": "Multi-channel Uplift Policy Learning", "published": "2026-07-30T13:19:35Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28135", "title": "LM-GRASP: Instance-Specific Language Models for Combinatorial Construction via Online Imitation Learning", "published": "2026-07-30T12:45:18Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28127", "title": "FinSMART: Financial Sentiment Analysis for Algorithmic Trading through Market-Aligned Reinforcement Learning", "published": "2026-07-30T12:36:32Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.28100", "title": "PCAP-LM: An LLM-Native Text Representation for TLS Bulk Traffic Analysis", "published": "2026-07-30T12:10:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28082", "title": "GGC: Selective Query Correction for Reliable Text-to-SPARQL Generation", "published": "2026-07-30T11:52:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28077", "title": "LEEPS: Latent-Guided Explore-Exploit Prompt Sampling for Efficient RLVR in Large Language Models", "published": "2026-07-30T11:50:10Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.28074", "title": "Echoverse: Deep, Evolving Environments for Training Computer-Use Agents at Scale", "published": "2026-07-30T11:48:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28073", "title": "GVR-Coder: A Visual-Feedback Framework for Structured SVG Generation in Complex Document and Meeting Scenarios", "published": "2026-07-30T11:47:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28048", "title": "SKILL-KD: Contrastive Skill Distillation for LLM Agents", "published": "2026-07-30T11:27:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28035", "title": "Enhancing Irregular Time Series Forecasting with Continuous-Time Modeling Framework", "published": "2026-07-30T11:18:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28026", "title": "Contrastive Reinforced Policy Optimization via Privileged Self-Distillation", "published": "2026-07-30T11:14:11Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "on-policy-distillation", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B09", "plan_reason": "Rubric、外部 rollout 与多奖励 RL — completed" }, { "arxiv_id": "2607.27968", "title": "Beyond Binary Rewards: A Comparative Study of Reward Design for Reinforcement Unlearning", "published": "2026-07-30T10:15:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27967", "title": "MARS-RA: Rank Aggregation for Credit Assignment via Multimodal Comparisons in Embodied Multi-Agent Cooperation", "published": "2026-07-30T10:14:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27944", "title": "Interpretable Representation via LLM-Driven Generative Disentanglement for Local-Life Service Recommendation", "published": "2026-07-30T09:52:18Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-kuaishou" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.27928", "title": "Harnessing the Potential of Optimizing Data Mixtures via Bayesian Domain Reweighting", "published": "2026-07-30T09:41:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27919", "title": "Memory Decoder at Scale: A Pretrained, Parametric Long-Term Memory", "published": "2026-07-30T09:30:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27914", "title": "Exact Action Values Are Not Enough: Rollout-Verified Reinforcement Fine-Tuning of a Reasoning Model for Multi-Zone VAV Control", "published": "2026-07-30T09:28:37Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27913", "title": "S-CEReBrO: Breaking the Memory Barrier in Continuous EEG Monitoring", "published": "2026-07-30T09:27:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27895", "title": "MMHBench: A Multi-Perspective Benchmark for Mental Health Understanding in Long-Form Videos", "published": "2026-07-30T09:12:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27881", "title": "RoboBRIDGE: A Modular Framework for Bridging Policies to Robust Real-World Robotic Agents", "published": "2026-07-30T08:55:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27879", "title": "ARES: Adaptive Reasoning-Effort Steering for PPA- and Cost-Aware RTL Optimization with LLM Agents", "published": "2026-07-30T08:53:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28692", "title": "SciToolAgent-Evo: An Ontology-Aware Self-Evolving Agent for Open-World Scientific Tool Acquisition", "published": "2026-07-30T08:47:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27853", "title": "FinanceHarness: Autonomous Financial Deep Research Framework", "published": "2026-07-30T08:32:04Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27844", "title": "Nanoparticle Networks for Neuromorphic Computing", "published": "2026-07-30T08:24:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27834", "title": "MemTxn: A Transaction Boundary for Source-Supported Updates and Complete-State Recovery in Agent Memory", "published": "2026-07-30T08:15:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27816", "title": "Beyond Borrowed Histories: Person-Aligned User Simulation for Interactive Role-Playing Evaluation", "published": "2026-07-30T07:55:35Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27797", "title": "Revisiting Predictive Process Monitoring in the Age of Foundation Models: A Comparative Study of Sequence, Tabular, and LLM Approaches", "published": "2026-07-30T07:36:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27789", "title": "From Understanding to Action: Feedback-Grounded Policy Discovery for Generative Recommendation", "published": "2026-07-30T07:23:07Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B02", "plan_reason": "7 月工业生成推荐、Agent harness 与搜索 — completed" }, { "arxiv_id": "2607.27773", "title": "ChronoMem: Version Control and Semantic Rollback for Large Language Model Agent Memory", "published": "2026-07-30T07:07:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27766", "title": "Gradient-free Task-Conditioned Retrieval for On-Device In-Context Learning", "published": "2026-07-30T07:03:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27760", "title": "Hierarchical Latent Reasoning for LLM-based Recommendation", "published": "2026-07-30T06:58:37Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-amazon" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.27756", "title": "Cocktail-Talker: Multi-Speaker Dialog Modeling in Noisy Social Environments with Turn Action GRPO", "published": "2026-07-30T06:53:15Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27739", "title": "Measuring Alignment With Reader Highlights Net of Position and Length", "published": "2026-07-30T06:24:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27735", "title": "A Sparse Glimpse of the Whole: Train-Free Self-Speculative Decoding", "published": "2026-07-30T06:16:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27705", "title": "Albilich: Steerable Proof-State Orchestration for LLM-Based Mathematical Research with CAS Integration", "published": "2026-07-30T05:41:44Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27704", "title": "LightRot: A Light-Weighted Rotation Scheme and Architecture for Accurate Low-Bit Large Language Model Inference", "published": "2026-07-30T05:39:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27703", "title": "SpatialCLI: Learning to Reason With Spatial Tools, Then Without Them", "published": "2026-07-30T05:39:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27699", "title": "RefineSVG: Visual Feedback-Driven Reinforcement Learning for Image-to-SVG Generation", "published": "2026-07-30T05:30:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27694", "title": "GyRot: Leveraging Hidden Synergy between Rotation and Fine-grained Group Quantization for Low-bit LLM Inference", "published": "2026-07-30T05:26:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27682", "title": "Restoring Collaborative Signals in Semantic-ID Generative Recommendation via Personalized Natural Language", "published": "2026-07-30T04:58:22Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27652", "title": "Harness-G: A Graph-Structured Harness for Search Agents", "published": "2026-07-30T04:05:05Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27647", "title": "LoopMemGR: From Behavior Logs to Evolving Memory for Generative Recommendation", "published": "2026-07-30T03:59:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "priority-org-taobao", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.27640", "title": "Dynamic Exploration Graph: A Novel Approach for Efficient Nearest Neighbor Search in Evolving Multimedia Datasets", "published": "2026-07-30T03:50:40Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27631", "title": "ReDiPPO: Reference-Guided Value Calibration and Discrepancy-Aware Token Reweighting for Mathematical Reasoning", "published": "2026-07-30T03:42:52Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27623", "title": "An Exploration Graph with Continuous Refinement for Efficient Multimedia Retrieval", "published": "2026-07-30T03:22:17Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27610", "title": "Kalman Meets Curriculum: Efficient Dynamic Prompt Selection for Adaptive RL Finetuning", "published": "2026-07-30T02:55:02Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation", "opd" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.27600", "title": "Back from the Future: Key-Value Cache Management by Counter-Causal Surprise", "published": "2026-07-30T02:42:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27599", "title": "World Action Planner: Generalizable Decision-Making with Action-Conditioned World Models", "published": "2026-07-30T02:41:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27597", "title": "A Systems Engineering Framework for Vision-Language-Enabled UAV Triage and Disaster Response", "published": "2026-07-30T02:35:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27595", "title": "Beyond Similarity: Grounded Agentic Extraction and Expert-Adjudicated Evaluation of Intertextuality in Classical Chinese Histories", "published": "2026-07-30T02:34:23Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27581", "title": "MUGEN: A Unified Framework for Efficient Motion Understanding and Generation", "published": "2026-07-30T01:55:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27574", "title": "Policy Gradient Steering: Interventions from Behavioral Objectives", "published": "2026-07-30T01:29:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27562", "title": "DeepResearch Agent System", "published": "2026-07-30T01:15:47Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agentic-rl", "multi-agent", "tool-agent", "web-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.27557", "title": "Training Skills Like Parameters via Self-Supervised Semantic Diffusion", "published": "2026-07-30T01:03:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27556", "title": "Evaluating Agentic Bioinformatics through Function, Evidence, and Validation", "published": "2026-07-30T01:00:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27553", "title": "AI and Its Impact on Creativity and Diversity: An Empirical Study of LLM-Generated Product Ideas", "published": "2026-07-30T00:55:25Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27528", "title": "ThreatForest: Multi-Agent Attack Tree Generation with Pluggable TTP Framework Mapping", "published": "2026-07-29T23:46:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27512", "title": "Belief Coevolution in a Social Network of Generalist and Specialist Large Language Models", "published": "2026-07-29T23:02:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27506", "title": "Models for minimalist RAG: B1ade 335M Embedding and 1B Parameter Small Language Models", "published": "2026-07-29T22:41:25Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27501", "title": "A Lightweight Foundation Model for Collider Physics with Multi-Domain Adaptation", "published": "2026-07-29T22:35:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27475", "title": "OneShot: Index-in-Ranking with Neural Scoring for Large-Scale Retrieval", "published": "2026-07-29T21:29:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "industrial-ranking", "recsys-general" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B00", "plan_reason": "PR #120 / OneShot" }, { "arxiv_id": "2607.27418", "title": "Context-Informed Ship Trajectory Prediction via Conditional Attention", "published": "2026-07-29T19:38:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27415", "title": "Bridging Inference-Time Scaling and Episodic Memory with Action-Centric Graphs", "published": "2026-07-29T19:31:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27372", "title": "Explorative Modeling: Unlocking a Third Pretraining Axis and End-to-End Generation", "published": "2026-07-29T18:25:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27366", "title": "BridgeAlign: Bridging Preference Alignment for Humanities and Social Sciences", "published": "2026-07-29T18:22:50Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27360", "title": "SkillMentor: LLM Agent Self-Evolution via Learning Blind-Spot Diagnosis", "published": "2026-07-29T18:13:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28680", "title": "TELLER: Dual-Path Iterative Preference Optimization for Table Entity Linking", "published": "2026-07-29T18:08:26Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "model-compression", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27201", "title": "Mental World Modeling", "published": "2026-07-29T17:59:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27178", "title": "DenseOn with the LateOn: Fully Open Dense and Late-Interaction Models for Multilingual, Long-Context, and Code Search", "published": "2026-07-29T17:50:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27303", "title": "THGFM: Dual-Branch Temporal Heterogeneous Graph Fusion Model", "published": "2026-07-29T17:04:30Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27132", "title": "Minimal Markovization via Stable Quotients in Holonomy-Cover Decision Processes", "published": "2026-07-29T17:03:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27130", "title": "AgentMap: Joint Equivalence and Subsumption Discovery for Ontology Matching", "published": "2026-07-29T16:58:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27090", "title": "InferScale: GPU-Native KV Injection for Personalized LLM Serving", "published": "2026-07-29T16:18:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27081", "title": "On-Policy Distillation for LLM Safety: A Routing Approach to Template-Robust Realignment", "published": "2026-07-29T16:07:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27056", "title": "Setoka: A Benchmark for Hierarchical User Understanding in Personalized Agents over Heterogeneous Data", "published": "2026-07-29T15:47:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27002", "title": "IMFuse: Instance-Aware Multi-Layer Fusion for LLM-Enhanced Sequential Recommendation", "published": "2026-07-29T14:58:36Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26998", "title": "AgentSnare: Learning to Delay, Divert, and Defuse Autonomous Penetration Agents", "published": "2026-07-29T14:56:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26993", "title": "Foundation Models for Face Presentation Attack Detection: A Unified Linear-Probing Benchmark", "published": "2026-07-29T14:52:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26981", "title": "OptimismBench: Forecasting Bias and the Alignment Effect in Language Model Judgment", "published": "2026-07-29T14:38:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26914", "title": "BioVLN: A Simulation Platform for Visual Language Navigation in Biomedical Laboratories", "published": "2026-07-29T13:44:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26908", "title": "Actions Have Consequences: Detecting Outcome Performativity using Intervention Testing", "published": "2026-07-29T13:39:29Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26893", "title": "Beyond Action Imitation: Learning a Decision-Aware User Simulator for Online Advertising", "published": "2026-07-29T13:25:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-tencent", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.26891", "title": "DIRECT: Direct Decoding for Efficient and Aligned Sequence Labeling with Large Language Models", "published": "2026-07-29T13:23:30Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26873", "title": "SERPO: Self-Evolving Rubric Policy Optimization for Open-Ended Test-Time Reinforcement Learning", "published": "2026-07-29T13:03:07Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "reward-model", "test-time-rl" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B09", "plan_reason": "Rubric、外部 rollout 与多奖励 RL — completed" }, { "arxiv_id": "2607.26832", "title": "Kairos: Numerically Robust News Recommendation under Item Cold-Start via Cholesky-based LinUCB", "published": "2026-07-29T12:25:25Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27273", "title": "SDO: Structure-Aware Data Organization for Efficient LLM Post-Training", "published": "2026-07-29T12:17:18Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26820", "title": "Forecasting Trajectory-Level Safety Risks in Black-Box Multi-Turn Interactions", "published": "2026-07-29T12:14:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27271", "title": "RLPF: Reinforcement Learning from Performance Feedback for Code Generation", "published": "2026-07-29T11:39:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26791", "title": "SecRespond: Benchmarking AI Agents for Real-World Post-Compromise Incident Response", "published": "2026-07-29T11:32:23Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26773", "title": "Do Latent Channels Actually Communicate? A Causal Audit of Latent Multi-Agent LLM", "published": "2026-07-29T11:14:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26760", "title": "Metis: Memory Foundation Model", "published": "2026-07-29T10:58:44Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "model-compression", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27269", "title": "Beyond KV Reconstruction: Functional Reconstruction for MLA Draft Models in Speculative Decoding", "published": "2026-07-29T10:54:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26724", "title": "UrbanDS: A Graph-Guided LLM Multi-Agent System for Data-Intensive Urban Tasks", "published": "2026-07-29T10:14:02Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agent-planning", "multi-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.26720", "title": "CaIRec: Calibrated Modality Imputation for Incomplete Multimodal Recommendation", "published": "2026-07-29T10:10:20Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26661", "title": "AgenticCANN: Automated Ascend C Operator Generation via Knowledge-Augmented Agentic Evolution", "published": "2026-07-29T09:19:17Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26656", "title": "Graph Is the Verifier: Agentic Reinforcement Learning for Interprocedural Vulnerability Detection", "published": "2026-07-29T09:16:54Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26654", "title": "Constitutional Midtraining: Content Presence Drives Alignment Gains", "published": "2026-07-29T09:15:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26648", "title": "The Sparsity Ceiling: Where Spiking Networks Can and Cannot Trade Activity for Energy", "published": "2026-07-29T09:09:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26642", "title": "AlphaSchema: Exploring the Space of Trading Semantics for LLM-Based Alpha Mining", "published": "2026-07-29T09:04:25Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26637", "title": "Filesystem-Based Memory for LLM Agents: Organization, Evolution, and Sustainability", "published": "2026-07-29T08:59:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26631", "title": "RAG-HAR+: Towards Cost-Efficient LLM-Based Human Activity Recognition for Edge Deployment", "published": "2026-07-29T08:57:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26621", "title": "WhisperRec: Latent Reasoning for Efficient Foundation Recommendation Models", "published": "2026-07-29T08:48:35Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-kuaishou" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.26604", "title": "WikiLoop: Jointly Learning to Build and Navigate Agent-Native Wikis with Downstream Feedback", "published": "2026-07-29T08:28:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26598", "title": "Living-Harness Is an Interactive-Agent Evolver", "published": "2026-07-29T08:20:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14644", "title": "DUET: Dual-Teacher On-Policy Distillation via Same-Weight Disagreement for Prohibition Compliance", "published": "2026-07-29T07:17:43Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "on-policy-distillation", "post-training", "preference-optimization" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.26523", "title": "The Art of Not Forgetting A Local Learning Architecture for Continual Learning", "published": "2026-07-29T06:42:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26515", "title": "HiFloat4 Format for End-To-End Reinforcement Learning Post-Training of Large Language Models", "published": "2026-07-29T06:28:26Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26500", "title": "Multi-Decoder OneRec: Controllable Generative Retrieval for Multi-Objective Industrial Recommendation", "published": "2026-07-29T05:54:19Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.26490", "title": "EvoPINN: Agentic Discovery of Executable Algorithms for Physics-Informed Neural Networks", "published": "2026-07-29T05:36:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26485", "title": "Parameterized Fair Resource Allocation under Diversity Constraints", "published": "2026-07-29T05:25:02Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28098", "title": "SciDataSailor: Deep Scientific Data Exploring", "published": "2026-07-29T05:08:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26473", "title": "Learning Dynamic User Personas from Implicit Interaction Streams via Iterative Refinement", "published": "2026-07-29T04:56:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26467", "title": "Neural Architecture Search for Traffic Prediction: A Survey of Methods, Challenges, and Future Directions", "published": "2026-07-29T04:41:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26457", "title": "DHRCL:Training Code LLMs with Dense Hierarchical Rewards and Curriculum Learning", "published": "2026-07-29T04:16:33Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.26448", "title": "Mergeable Model-Side Aggregation States for Long-Context Language Models", "published": "2026-07-29T03:56:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26429", "title": "NMKFR: A Robust Framework for Time-Aware Cold-Start Recommendation", "published": "2026-07-29T03:15:36Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26427", "title": "PSG: Pair-Space Generation for Efficient Generative Reranking", "published": "2026-07-29T03:14:09Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "priority-org-kuaishou", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.27258", "title": "PlantBGC: Transformer for Plant BGC Discovery via Label-Free Domain Adaptation and Weak Supervision", "published": "2026-07-29T03:13:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26418", "title": "DIRECTOR: Dynamic Index-based Recommendation with Transport-Optimized Retrieval", "published": "2026-07-29T03:03:03Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.26393", "title": "CaM-Wolf: Causal-Aware Multimodal Agents for Social Deduction Games", "published": "2026-07-29T01:58:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26391", "title": "Q-Steer: Action-Value Guidance for Molecular Policy Optimization", "published": "2026-07-29T01:53:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26380", "title": "Continuous Online Evaluation of Recommendation Strategies in Social Science Academic Search", "published": "2026-07-29T01:33:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26369", "title": "ClockRoPE: Random Fourier Rotations for Temporal Routine Modeling", "published": "2026-07-29T01:07:27Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "generative-recommendation", "production-evidence", "recsys-general" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B00", "plan_reason": "PR #120 / ClockRoPE" }, { "arxiv_id": "2607.26365", "title": "Embedding Items at Scale: Comparing GNN-Based and ID-Based Item Embeddings in the Yandex Ecosystem", "published": "2026-07-29T00:44:25Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20402", "title": "LingShu: A Large-Scale Symptom-Centric Contextualized Knowledge Graph Bridging Traditional Chinese Medicine and Modern Biomedicine", "published": "2026-07-29T00:36:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26358", "title": "Post-Training at the Edge of Detectability: A Game-Theoretic Approach to Fine-Tuning", "published": "2026-07-29T00:19:09Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "post-training", "reward-model", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.14642", "title": "Training and Evaluating Ethical Reinforcement Learning Agents on Per-Episode Distributions", "published": "2026-07-29T00:00:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26350", "title": "Dissecting Sensitivity to Training Language in Self-Supervised Speech Learning Using Neural Audio Codec Tokens", "published": "2026-07-28T23:46:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26336", "title": "Learning Implicit Causal World Models from Multi-Agent Demonstrations", "published": "2026-07-28T23:18:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26298", "title": "Entity Resolution in Practice: Lessons from a Self-Serve Pipeline", "published": "2026-07-28T21:41:32Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26220", "title": "Model-Driven Requirements Configuration with Three-Valued Uncertainty Scoring", "published": "2026-07-28T19:49:42Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26212", "title": "Multi-Agent Debate Strategies: Survey, Taxonomy, and Challenges", "published": "2026-07-28T19:26:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26179", "title": "Cognitive Convergence: Deep Similarities Between Large Language Models and Human Cognition", "published": "2026-07-28T18:37:48Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26120", "title": "Even More Deception: Objective Misalignment in Mixed-Motive LLM Multi-Agent Systems", "published": "2026-07-28T17:48:54Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26119", "title": "Probing the Origins of Reasoning Performance: Representational Quality for Mathematical Problem-Solving in RL vs. SFT Fine-Tuned Models", "published": "2026-07-28T17:42:42Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26015", "title": "Instruction-Tuned Models Locally Reuse Human Syntax More Than Humans Do", "published": "2026-07-28T17:27:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25988", "title": "Generator-Aligned Representation Interfaces for Diagnostic Soft Equivariance", "published": "2026-07-28T17:06:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25970", "title": "Reinforcement Learning for Code Optimization", "published": "2026-07-28T16:52:31Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25947", "title": "A Cost-Effective Multimodal LLM Reasoning Framework for Question Answering over Irregular Clinical Time Series", "published": "2026-07-28T16:33:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25921", "title": "Evaluating VLMs for Autonomous Agent-Driven Geometry Clipping Detection in Video Game QA", "published": "2026-07-28T16:10:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26115", "title": "GPT-Red: Automated Red Teaming via Self-Play at Scale", "published": "2026-07-28T16:03:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25904", "title": "Interactive Reward Agent: GUI Task Evaluation via Environment-State Verification", "published": "2026-07-28T16:01:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25895", "title": "HiFi-UMI: Learning Deployable Manipulation Policies from High-Fidelity UMI Data Alone", "published": "2026-07-28T15:52:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25886", "title": "RSIBench-Data: Benchmarking Data-Centric Research for Recursive Self-Improvement", "published": "2026-07-28T15:46:41Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25877", "title": "Runtime Uncertainty Monitoring for LLM-Based Multi-Agent Systems Using Bayesian Networks", "published": "2026-07-28T15:39:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25823", "title": "Hypothesis-Driven Shelf Generation for Personalised Recommendation", "published": "2026-07-28T15:05:42Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.25816", "title": "Speculate While You Reason: Teaching Agents to Predict Their Next Tool Call via Joint Agent-Speculator RL", "published": "2026-07-28T15:00:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25765", "title": "WorkSurface-Bench: Benchmarking Enterprise Agents on Multi-Surface Knowledge Routing", "published": "2026-07-28T14:19:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.27251", "title": "Recursive transformers for semiconductor thermo-mechanical reliability", "published": "2026-07-28T13:58:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25642", "title": "Instruction-based Image Editing: A Survey on Data, Models, Evaluation, and Applications", "published": "2026-07-28T12:27:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25640", "title": "LLM-as-a-Judge for Evaluating System Responses in Conversational Music Recommendation", "published": "2026-07-28T12:26:55Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.25485", "title": "PatientAgentBench: A Benchmark Framework for Evaluating Patient-Facing Health AI Agents", "published": "2026-07-28T09:24:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25479", "title": "Architectural Backdoors in Vision-Language Model Supply Chains via Representation Steering", "published": "2026-07-28T09:12:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14637", "title": "Early Cycle Charge Trajectory Generative Prediction and Full Life Cycle Health Management of Iron-Chromium Flow Batteries Based on FlowBD-E1", "published": "2026-07-28T09:02:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25459", "title": "Emergent Latent-State Computation under Stochastic Volatility", "published": "2026-07-28T08:49:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25451", "title": "Bits and Memories: Measuring Verbatim Extraction Across LLM Quantization", "published": "2026-07-28T08:41:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25446", "title": "Toward an Organizational Science of Multi-Agent LLM Systems: Decoupling Who, How, and Which Algorithm", "published": "2026-07-28T08:35:21Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "multi-agent", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25420", "title": "MARS: Multi-Agent Re-ranking for Repeat-Order Food Delivery Recommendation", "published": "2026-07-28T08:18:08Z", "tracks": [ "agent", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "multi-agent", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25415", "title": "A Control System, a Dataset, and a Recipe for Making Frozen LLM Agents Learn a Domain", "published": "2026-07-28T08:10:25Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25408", "title": "Context Assembly as the Controlled Variable: A Control-Theoretic View of Harness Policies for Frozen LLM Agents", "published": "2026-07-28T08:07:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25400", "title": "COVENANT: Natural-Language Workflow Compilation for Aligned Agent Execution", "published": "2026-07-28T07:59:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25398", "title": "HANDBOOK.md: A Benchmark for Long-Context Agentic Instruction Following", "published": "2026-07-28T07:58:07Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "long-context", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09967", "title": "SPOTting the Future: Lookahead Explanations for Deep Reinforcement Learning", "published": "2026-07-28T07:57:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25379", "title": "Cyber-Capable AI Agents: Vulnerabilities, Evaluation Containment, and Defensive Response", "published": "2026-07-28T07:34:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25366", "title": "Sharpness-aware Model Merging with Salience Recovery for LLM-based Cross-Domain Sequential Recommendation", "published": "2026-07-28T07:17:44Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25357", "title": "Raven: High-Recall Sequence Modeling with Sparse Memory Routing", "published": "2026-07-28T07:04:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25346", "title": "The Case Against Generation for Retrieval: Discriminative Language Models as Effective Retrievers", "published": "2026-07-28T06:47:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "deployment-evidence", "generative-recommendation", "llm-recommendation", "recommendation-ranking", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.25344", "title": "Reward Guided Decoding for Generative Recommendation", "published": "2026-07-28T06:46:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "generative-recommendation", "priority-org-kuaishou" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.25340", "title": "Cardiologent: Multi-Agent Clinical Decision Support for Patient-Level Arrhythmia Assessment, Urgency, and Management", "published": "2026-07-28T06:43:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25339", "title": "SPARC: Sequence-aware Progressive Attribute Routing and Compression Framework for Generative Recommendation", "published": "2026-07-28T06:43:25Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "priority-org-amazon", "priority-org-taobao" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.25337", "title": "Temporal-Distance JEPA: Plan-Aware Representation Learning for Latent World Model Predictive Control", "published": "2026-07-28T06:38:16Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25329", "title": "Grevo: A Unified Generative Recommendation Framework with Evolutionary Item Indexing", "published": "2026-07-28T06:27:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26099", "title": "Lilith: Backdoor Generalization under Training-Inference Trigger Shift", "published": "2026-07-28T06:23:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25297", "title": "Hybrid Analysis for Secure MCP Tool Use in LLM Agents", "published": "2026-07-28T05:17:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25294", "title": "CLBench-V: Evaluating Multimodal Context Learning from Grounding to Knowledge Acquisition", "published": "2026-07-28T05:06:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25291", "title": "CoSA: Accelerating Long-Context Inference via Proxy-Kernel Co-Designed Sparse Attention", "published": "2026-07-28T04:57:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25276", "title": "FunnelAL: Retrieve-then-Rank Active Learning for Single-Class Discovery", "published": "2026-07-28T04:26:26Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "industrial-ranking", "recommendation-ranking", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.25257", "title": "Laplace-PSN-IRT: Uncertainty Quantification for Neural Item Response Theory Models of LLM Benchmarks", "published": "2026-07-28T03:56:23Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25253", "title": "The User Asks, Platforms Compete: How Agentic Recommendation Markets Take Shape", "published": "2026-07-28T03:54:38Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "recommendation-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18132", "title": "Alignment Is All You Need: Instruction-Free Training for General Audio-Language Models", "published": "2026-07-28T03:30:41Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training", "preference-optimization", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25227", "title": "Decision-Level Hijacking: Injecting Cognitive Bias into Large Language Models via Bit-Flip Attacks", "published": "2026-07-28T02:59:32Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25216", "title": "TopoGR: Revealing and Preserving Latent Structure of Semantic ID in Generative Recommendation", "published": "2026-07-28T02:45:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25209", "title": "VaLiDRec: Variable-Length LLM-Aligned Semantic IDs for Generative Recommendation", "published": "2026-07-28T02:29:10Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25207", "title": "A Unified Algorithmic Framework for Hybrid Reinforcement Learning in Tabular MDPs with Shifted Transition Dynamics", "published": "2026-07-28T02:27:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26094", "title": "Meta-Learned Reward Shaping for Reinforcement Learning from Human Feedback", "published": "2026-07-28T02:09:26Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "preference-optimization", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.14635", "title": "Belayer: Efficient Fault Tolerance for LLM Agentic RL Training", "published": "2026-07-28T01:46:18Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "efficient-inference", "llm-rl", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25145", "title": "Agentic AI for Scientific Reasoning in Autonomous Quantum Sensing Experiments", "published": "2026-07-27T23:43:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25140", "title": "How Affect Propagates among LLM Agents: Emergent Emotional Contagion in Crowd Simulation", "published": "2026-07-27T23:22:17Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25110", "title": "Memory Layer: Train the In-Model Cache for Recommendation Models", "published": "2026-07-27T22:15:13Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.25097", "title": "On the Convergent Validity of Offline Evaluation Designs for Recommender Systems", "published": "2026-07-27T21:54:19Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25091", "title": "Towards Robust Reinforcement Learning for Small-Scale Language Model Agents", "published": "2026-07-27T21:30:44Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agentic-rl", "llm-rl", "preference-optimization", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.25069", "title": "DS@GT ARC at CheckThat! 2026: LLM-Based Trace Ranking and Grouped Reward Modeling for Multilingual Numerical Claim Verification", "published": "2026-07-27T20:59:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25066", "title": "Addressable Recall Compaction for Long Context-Window Control in AI Agents", "published": "2026-07-27T20:51:05Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25041", "title": "ScoreShield: Differentially Private Release of Similarity Scores", "published": "2026-07-27T20:02:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.25032", "title": "Authoring Agent Skills: A Software-Engineering Approach", "published": "2026-07-27T19:43:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24743", "title": "ClinFusion: A Vision-Centric Multimodal LLM System for Holistic Medical Understanding", "published": "2026-07-27T17:59:49Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24720", "title": "The Physics of Multi-Turn Long-Horizon Planning: From Pre-training to Post-training via Single- and Multi-Teacher On-Policy Agentic Distillation", "published": "2026-07-27T17:55:03Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "on-policy-distillation", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24688", "title": "Beyond Scale and Generation: Understanding Language Model-based Entity Matching", "published": "2026-07-27T17:29:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24665", "title": "MMOE: Modernizing Diffusion Transformers with Efficient Expert Design", "published": "2026-07-27T17:05:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24653", "title": "Kimi K3: Open Frontier Intelligence", "published": "2026-07-27T16:49:54Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-rl", "long-context", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24651", "title": "Evidence Attribution in Visual Document Understanding without Coordinates or Region Labels", "published": "2026-07-27T16:49:36Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24900", "title": "Inverse RL Helps Align AI by Imitating Humans", "published": "2026-07-27T16:45:31Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "preference-optimization", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.24617", "title": "LaRec: Unleashing LLM-based Latent Reasoning for Generative Recommendation", "published": "2026-07-27T16:14:57Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24607", "title": "One Graph, Multiple Gains: Single High-Quality Item-Item Graph for Multimodal Recommendation", "published": "2026-07-27T16:09:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24588", "title": "SIREN: Towards End-to-End Extreme-Weather Early Warning with Experience-Grounded LLM Agents", "published": "2026-07-27T15:53:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24555", "title": "LOCKS: Page-Local Compact Key Summaries for Efficient Long-Context Decoding", "published": "2026-07-27T15:28:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24542", "title": "From transcription to semantic corpus analysis: unsupervised learning of sentence representations for ancient languages", "published": "2026-07-27T15:20:13Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24532", "title": "From Machine Learning to Large-Scale EO Products: Best Practices for Making Maps", "published": "2026-07-27T15:10:57Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24522", "title": "FlowCTS: On-policy Continuous Trajectory Supervision of Flow Models", "published": "2026-07-27T15:03:22Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "on-policy-distillation", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.24513", "title": "Physics Transformer: Tailoring Transformer for General PDE Prediction", "published": "2026-07-27T14:53:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24484", "title": "What do Reward Models Memorize?", "published": "2026-07-27T14:20:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24471", "title": "Grounding latent algorithm routing in transformer reasoning", "published": "2026-07-27T14:07:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24449", "title": "Evaluating RAG for French immigration law: a benchmark and baseline study", "published": "2026-07-27T13:55:35Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24440", "title": "Bigger or Cheaper? Scale and Quantization Effects on Uncertainty Signals in Vision-Language Models Under Image Degradation", "published": "2026-07-27T13:47:44Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24430", "title": "Let Me Look at You: Advanced Facial Expression Modeling for Conversational Speech Synthesis", "published": "2026-07-27T13:42:37Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24402", "title": "CogRec: Structure-Cognitive Fast-and-Slow Reasoning for Generative Recommendation", "published": "2026-07-27T13:19:55Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24392", "title": "When LLM Defenses Backfire: Characterizing Safety, Performance, and Cost Trade-offs", "published": "2026-07-27T13:07:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24368", "title": "Keep It InMind: Benchmarking the Implicit-Association Blind Spot in Agent Memory", "published": "2026-07-27T12:42:12Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "agent-memory", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24352", "title": "Retrieval-Augmented Large Language Models as Components of Cognitive Computing architecture for Regulatory Knowledge Management", "published": "2026-07-27T12:31:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24341", "title": "Simulating Tenant Responses to Energy Policy Interventions with Transaction-Cost-Aware LLM Agent", "published": "2026-07-27T12:18:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24331", "title": "DynaCalKV: Key-Value Cache Compression via Head Grouping and Adaptive Rank Allocation", "published": "2026-07-27T12:08:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24312", "title": "CONSISTRE: A Unified Consistency-Aware Framework for Document-Level Relation Extraction with Large Language Models", "published": "2026-07-27T11:56:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24280", "title": "From Proprietary to Open-Source: Bridging the Distribution Gap via Multi-Agent Protocol Distillation in Agentic Search", "published": "2026-07-27T11:27:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24273", "title": "INS-ActBench: A Comprehensive Benchmark for Assessing Professional Actuarial Capability of Large Language Models", "published": "2026-07-27T11:12:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24241", "title": "FilmBench: A Film-Grade Benchmark for Cinematic Video Generation", "published": "2026-07-27T10:20:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24213", "title": "Integrating Factual and Normative Industrial Knowledge via Constraint-Aware Graph Attention for Process Plan Recommendation", "published": "2026-07-27T09:45:05Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "industrial-ranking", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.24192", "title": "LLM-based Source Code Compression via Thresholded Symbol Ranking", "published": "2026-07-27T09:12:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24092", "title": "ConAlign: Conditional Alignment Framework for Balancing Biased and Unbiased Recommendation", "published": "2026-07-27T07:28:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-kuaishou", "production-evidence", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.24062", "title": "ACRL: Adaptive Control of Training-Inference Discrepancy for Stable Reinforcement Learning", "published": "2026-07-27T07:05:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24051", "title": "HELIOS: An LLM-Driven Autonomous Indirect Trajectory Optimization Agent", "published": "2026-07-27T06:48:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24025", "title": "SpecFormer: Mitigating Embedding and Attention Collapse via Spectral-Aware Transformer for Recommendation", "published": "2026-07-27T05:53:00Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "industrial-fulltext-review", "p2-deferred-review" ], "matched_queries": [ "industrial-ranking", "recsys-general", "transformer-architecture" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.23991", "title": "SyRuP: Enhancing System-Prompt Following via Reward-Guided Prediction in LLM Decoding", "published": "2026-07-27T04:39:13Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23986", "title": "MEMOIR: Temporal Behavioral Memory for Recommendation Across the Preference-Drift Spectrum", "published": "2026-07-27T04:27:58Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-amazon" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.23933", "title": "SpecBox: Speculative Sandbox Scheduling for Efficient LLM Agent Serving", "published": "2026-07-27T02:10:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23909", "title": "WorldDiT: A Unified Diffusion Architecture for World and Action Modeling", "published": "2026-07-27T00:55:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24869", "title": "Ranked by Position: Order Sensitivity as an Exploitable Attack Surface in LLM Listwise Recommenders", "published": "2026-07-26T22:32:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-recommendation", "priority-org-amazon", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.23870", "title": "MulRobBench: A Decision-Level Benchmark for Safe and Security-Policy-Compliant Multimodal UAV Agents", "published": "2026-07-26T22:23:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23860", "title": "Controllable Diversity in Normalization-Based Implicit Ensembles via Softmax-Temperature Modulation", "published": "2026-07-26T22:06:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23811", "title": "Memory Efficient Audio Synthesis with Decoupled Temporal Depth Diffusion Transformers", "published": "2026-07-26T19:20:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23779", "title": "ClawRec: A Claw-Native Recommender System", "published": "2026-07-26T17:56:49Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23770", "title": "A Comparison of Data Augmentation Methods for Training Deep Neural Networks on Synthetic Aperture Sonar", "published": "2026-07-26T17:42:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23762", "title": "Escaping the Euclidean Void: Manifold-Informed Flow Matching for Sequential Recommendation", "published": "2026-07-26T17:17:31Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23753", "title": "On the post-hoc Evaluation of PDE Discovery: A Multifaceted Challenge of Scientific Advancement", "published": "2026-07-26T16:58:42Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23740", "title": "ZenGen: Social Mind for LLMs", "published": "2026-07-26T16:25:45Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "long-context", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23722", "title": "E-Bench: Benchmarking Multi-Step Tool-Use Agents in Real-World Product Scenarios", "published": "2026-07-26T15:38:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23678", "title": "Focus Is All You Need: Adaptive Goal-aware Attention Orchestration for Multi-Agent Graph Systems", "published": "2026-07-26T14:23:33Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-planning", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.24862", "title": "KuaiLive-M3: A Multi-Modal, Multi-Domain, and Multi-Feedback Dataset for Live Streaming Recommendation", "published": "2026-07-26T12:52:03Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-kuaishou" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23605", "title": "Hybrid Advantage Estimation with Unified Critic for VLM Agentic Reinforcement Learning", "published": "2026-07-26T11:16:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23561", "title": "Towards a Relevance Posterior in Neural Information Access", "published": "2026-07-26T09:24:20Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recommendation-ranking", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.23554", "title": "Neonatal Hypoxic-ischaemic Encephalopathy Classification from the EEG and HRV Signals Using a Conformer based Masked Autoencoder", "published": "2026-07-26T09:11:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23507", "title": "Choosing a Text Embedding Model: A Practical Benchmarking and Decision Framework", "published": "2026-07-26T07:13:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23491", "title": "PlanCraft: Sketch, Refine, and Furnish for Architect-Inspired Progressive 3D Residential Scene Generation", "published": "2026-07-26T06:35:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23488", "title": "Learning Sampling Parameters for Diffusion Models", "published": "2026-07-26T06:29:48Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "preference-optimization", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.23420", "title": "LA-RL: Label-Aware Self-Reflection for Reinforcement Learning in Information Extraction", "published": "2026-07-26T02:35:47Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.00050", "title": "Identifiability-Aware Source Apportionment in City-Scale Advection-Diffusion Systems", "published": "2026-07-25T23:10:51Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23364", "title": "On the Impossibility of Unbiased and Length-Invariant Policy Optimization with Outcome Rewards", "published": "2026-07-25T21:12:51Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23346", "title": "SPRKD: Effective Knowledge Distillation for Deep Neural Networks via Saddle Region Approximation", "published": "2026-07-25T19:53:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23175", "title": "Beyond a Global Norm: Personalizing Toxicity Sensitivity in Language Models Without Retraining", "published": "2026-07-25T12:09:43Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23153", "title": "In-Context Learning as Implicit Policy Gradient", "published": "2026-07-25T11:13:52Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23147", "title": "False Prophets: On the Security of World Models in Agentic Systems", "published": "2026-07-25T10:55:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23146", "title": "Foundation Models and Fine-Tuning: Toward a New Generation of Models for Time Series Forecasting", "published": "2026-07-25T10:55:06Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23125", "title": "Self-Boosting Vision-Language Models with Noisy Student On-Policy Self-Distillation", "published": "2026-07-25T10:00:33Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation", "post-training" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.23124", "title": "AgentOmnia: Scaling Agentic Models for Full-Scenario Applications", "published": "2026-07-25T09:58:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23123", "title": "SQBench: A Benchmark for Evaluating Task Delivery by Language-Model Agents in Production-Oriented Workflows", "published": "2026-07-25T09:55:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11224", "title": "Harnessing agent memory to build lifelong AI partners for materials scientists", "published": "2026-07-25T09:11:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23075", "title": "Traceable LLM Reasoning for Fake-Order Fraud Detection", "published": "2026-07-25T06:51:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24850", "title": "SearchArt: Training Long-Horizon Search Agent with Scalable Synthetic and Verified Task", "published": "2026-07-25T06:47:46Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23068", "title": "Neural Network-Driven Volatility Drag Mitigation under Aggressive Leverage", "published": "2026-07-25T06:32:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23054", "title": "Through the Bottleneck: How Multi-head Latent Attention Separates Content from Position in Language Models", "published": "2026-07-25T05:45:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23047", "title": "MixQuant: Adaptive Mixed-Precision Quantization for Large Language Models", "published": "2026-07-25T05:10:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23045", "title": "Stress-testing large language model agents in a robotic chemistry laboratory", "published": "2026-07-25T05:07:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.23038", "title": "EGR: Embedding-Native Generative Retrieval with a Shared LLM", "published": "2026-07-25T04:45:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-amazon", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.22999", "title": "WCM: World-Cognition Model for Generalizable Human-Robot Interaction", "published": "2026-07-25T02:27:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22997", "title": "Real2Sim2Real for Vision-Language-Action Manipulation: An AMD ROCm-Based Pipeline", "published": "2026-07-25T02:23:34Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22996", "title": "Beyond Direct Answering: Aligning Educational LLMs as Socratic Guides via Heuristic Reinforcement Learning", "published": "2026-07-25T02:21:32Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22962", "title": "ConsistencyGate: Preventing Memory Contamination in LLM Agents via Self-Consistency Admission Control", "published": "2026-07-25T00:08:09Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "agent-memory", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24846", "title": "Two Views, One Voice: Evidence-Grounded Conversational Music Recommendation", "published": "2026-07-24T23:12:02Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "industrial-ranking", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.24845", "title": "REPREC: Representation Driven Parameter-Efficient Recommendation System", "published": "2026-07-24T22:45:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22935", "title": "Discrepancy-Rounded Fair Bandits with Static and Time-Varying Exposure Floors", "published": "2026-07-24T22:20:44Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.02627", "title": "Micro-Segmentation Anomaly Detection in Zero-Trust Software-Defined Network Fabrics", "published": "2026-07-24T22:16:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22925", "title": "Not All LLM Reasoning is Visible in the Chain-of-Thought", "published": "2026-07-24T21:32:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22884", "title": "CHiPS: Character Histograms and Positional Signals for Lightweight Authorship Attribution in Romanian Texts", "published": "2026-07-24T19:48:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07532", "title": "Dynamic Coalition Formation and Communication Pricing in Skill-Based Agentic AI Systems", "published": "2026-07-24T19:25:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22837", "title": "Frustratingly Simple Black-Box Adaptation of Language Models via Logit Bias", "published": "2026-07-24T18:27:58Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22811", "title": "From Hybrid Mechanistic--Data-Driven Modeling Toward Neuro-Symbolic AI: What, Why, and How", "published": "2026-07-24T18:00:00Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22529", "title": "Skill Self-Play: Pushing the Frontier of LLM Capability with Co-Evolving Skills", "published": "2026-07-24T17:59:22Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22807", "title": "The Best Programming Language for Tokenmaxxing: An Investigation of Coding Agent Behavior Across Programming Languages", "published": "2026-07-24T16:41:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22465", "title": "TRACE-ROUTER: Task-Consistent and Adaptive Online Routing for Agentic AI", "published": "2026-07-24T16:29:06Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07531", "title": "Search-G1: Grounded Search Agents via Representation-Based Intrinsic Rewards", "published": "2026-07-24T15:22:57Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.22400", "title": "A Self-Calibrating Agentic AI Framework for Autonomous Edge Resource Allocation", "published": "2026-07-24T15:21:54Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22393", "title": "SceneActBench: Can Agents Act on the 3D Scenes They See?", "published": "2026-07-24T15:16:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22389", "title": "HiKV: Hierarchical Importance-Aware KV Cache with Hardware Acceleration for LLM Decoding", "published": "2026-07-24T15:15:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22375", "title": "IDEAgent: Agentic Quality-Diversity Search for Research Idea Generation", "published": "2026-07-24T15:03:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22361", "title": "Indexing: the Beginning and the End", "published": "2026-07-24T14:42:44Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22356", "title": "Integrated Order Dispatching and Routing for Last-Mile Pickup via Deep Reinforcement Learning", "published": "2026-07-24T14:37:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22351", "title": "IQ-JEPA: A Joint-Embedding Predictive Architecture with a Hermitian Vision Transformer for Sound Speed and Attenuation Estimation from Ultrasound IQ Data", "published": "2026-07-24T14:31:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22341", "title": "Bringing GRACE to Recommendation: Fine-Tuning for Sustainable and Accurate Personalization", "published": "2026-07-24T14:17:37Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22334", "title": "Cross-Tokenizer On-Policy Distillation via Byte-Prefix Marginalization", "published": "2026-07-24T14:12:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22319", "title": "Towards Trustworthy and Cost-Efficient Data Integration: From Naïve RAG to Agentic RAG", "published": "2026-07-24T13:58:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22314", "title": "Evolution-Aware MSA Reasoning for Subsampling via Factor Graphs", "published": "2026-07-24T13:55:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22287", "title": "Efficient Recommendations via Graph Coarsening and Label Propagation", "published": "2026-07-24T13:25:26Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22786", "title": "Optimizing Transformer Neural Network for Real-Time Outlier Detection on FPGAs", "published": "2026-07-24T12:09:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22785", "title": "FusionML: Prefill, Not Decode - Mechanism and Boundaries of CPU+GPU Co-Execution on Unified-Memory Apple Silicon", "published": "2026-07-24T11:42:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24841", "title": "Neuromorphic Diffusion Language Models: Addressing Compute and Memory Bottlenecks via Sparsity and Block Denoising", "published": "2026-07-24T11:14:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22186", "title": "Deconstructing Off-Policy Ratios: Entropy-Scaled Trust Regions for Asynchronous Reinforcement Learning", "published": "2026-07-24T10:55:42Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22157", "title": "Learning on the Job: Continual Learning from Deployment Feedback for Frozen-Weights Agents", "published": "2026-07-24T10:01:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22153", "title": "Industrial Tokenization for LLM-Based Health Intelligence: A Federated Architecture for Industrial Evidence Integration", "published": "2026-07-24T09:56:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22114", "title": "Pretraining EHR Foundation Models with Patient-Aware Sampling", "published": "2026-07-24T09:10:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22779", "title": "Multimodal Surface EMG Hand Gesture Recognition Using Query-Based Transformers for Prosthetic Control", "published": "2026-07-24T09:03:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22777", "title": "LC-SEPLM: long-range contact-supervised adaptation for sequence-only protein representation learning", "published": "2026-07-24T08:35:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22083", "title": "Nanbeige4.2-3B: Unlocking Agentic Capabilities in a Compact Model", "published": "2026-07-24T08:33:26Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data", "process-reward", "reward-model", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22776", "title": "Predicting the Outcome of rTMS Depression Therapy using EEG Signals and CNN", "published": "2026-07-24T08:23:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22068", "title": "Rethinking Multi-Branch and Cross-Backbone Fusion for Vehicle Re-Identification in the Foundation-Model Era", "published": "2026-07-24T08:10:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22067", "title": "Multimodal Language Models Benchmarked Against the NRC Reactor Operator Licensing Examination: Fine-Tuning and Retrieval Strategies", "published": "2026-07-24T08:10:21Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22043", "title": "Scaling Native Multimodal Pre-Training From Scratch", "published": "2026-07-24T07:13:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22039", "title": "Enough is as good as a feast: A Comprehensive Analysis of How Reinforcement Learning Mitigates Task Conflicts in LLMs", "published": "2026-07-24T07:08:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22035", "title": "DCS: A Unified Conditional Sensitivity Framework for Cross-Modal Copyright Infringement Detection", "published": "2026-07-24T07:00:44Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22034", "title": "Small Vision-Language Models Know When They Are Wrong But Cannot Say So: A Two-Model Study of Stated versus Internal Confidence Under Realistic Image Degradation", "published": "2026-07-24T07:00:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14632", "title": "DeMTS: Denoising Trajectories as Multivariate Time Series for Hallucination Detection in Diffusion Language Models", "published": "2026-07-24T06:45:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22014", "title": "Zero-Shot Mission-Level Evaluation for Aerial MLLM Agents", "published": "2026-07-24T06:22:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22012", "title": "Cross-Domain Off-Policy Evaluation and Learning for Contextual Bandits", "published": "2026-07-24T06:17:31Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21985", "title": "Unified Static-Dynamic Pruning for Efficient LLM Inference", "published": "2026-07-24T05:19:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21971", "title": "Teaching LLMs to Self-Evolve: Cultivating Core Meta-Skills with Reinforcement Learning", "published": "2026-07-24T04:35:29Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-rl", "post-training", "reward-model", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.22770", "title": "Dementia Etiology Diagnosis via Collaborative Meta Knowledge Enhancement", "published": "2026-07-24T04:26:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21958", "title": "Efficient Online LLM Watermark Detection via Rao-Blackwellized E-Processes", "published": "2026-07-24T04:10:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21951", "title": "SIREN (Luring LLMs onto the Rocks): PAIR-Driven Preference Manipulation in Web-RAG Recommenders", "published": "2026-07-24T03:55:51Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22769", "title": "DomainPilot: Domain-Level Loss-Guided Two-Stage Data Mixture Optimization for Efficient Language Model Fine-Tuning", "published": "2026-07-24T03:54:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22766", "title": "Beyond Shapley: An Influence-Based Data Auditing Pipeline for LLM Alignment and Evaluation", "published": "2026-07-24T03:31:44Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21927", "title": "RIS-Kernel: A Model-Agnostic Architecture for Long-Context LLM Inference via Sparse Attention", "published": "2026-07-24T03:00:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14629", "title": "Inference-Time Mitigation of Adversarial Political Bias in Large Language Models", "published": "2026-07-24T01:52:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21887", "title": "Towards Reducing Foreign Language Anxiety Using Level-Appropriate Embodied Conversational Agents", "published": "2026-07-24T01:31:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22763", "title": "An Integrated Deep Learning and Statistical Framework for Whole-Network Gene--Environment Association with Leaf Vascular Architecture", "published": "2026-07-23T23:23:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21850", "title": "SCALE: Self-Supervised Constraint-Aware Layout GEneration for Local P&R DRV Fixing at Advanced Nodes", "published": "2026-07-23T22:38:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21802", "title": "StARS: Socially Appropriate Robot Actions via a Recommender System-Driven Approach", "published": "2026-07-23T20:38:07Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22757", "title": "Hierarchical Grading in Large Language Models", "published": "2026-07-23T20:14:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "transformer-architecture" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21763", "title": "Every Model Cheats: Prompt-Level Mitigation of Cheating on Offensive Cyber Tasks", "published": "2026-07-23T19:26:25Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21574", "title": "Surprisal Theory is Tautological (without Rational Grounding)", "published": "2026-07-23T17:54:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21557", "title": "OpenForgeRL: Train Harness-native Agents in Any Environment", "published": "2026-07-23T17:38:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21552", "title": "MIRROR: Learning from the Other View for Multi-Modal Reasoning", "published": "2026-07-23T17:35:56Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21550", "title": "X$^3$-OPD: Distilling Reasoning into Large Audio-Language Models via On-Policy Alignment", "published": "2026-07-23T17:35:20Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21692", "title": "Learning What Matters: Supervising Global Context Pruning with Causal Evidence Sets", "published": "2026-07-23T17:11:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21522", "title": "GS-Agent: Creating 4D Physical Worlds With Generative Simulation", "published": "2026-07-23T17:04:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21519", "title": "Diffusion Language Model for Recommendation", "published": "2026-07-23T17:02:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.21513", "title": "Transparent by Design, Usable in Practice? A Formative Usability Study of a Conversational Product Advisor", "published": "2026-07-23T16:58:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21498", "title": "Artificial Epanorthosis: Why large language models overuse a classical rhetorical figure, and how to mitigate it", "published": "2026-07-23T16:47:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21488", "title": "Compact Latent Coordination for Autonomous Vehicles at Unsignalized Intersections", "published": "2026-07-23T16:28:35Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21482", "title": "Agentic coding without the cloud: evaluating open-weight large language models on longitudinal data preparation tasks", "published": "2026-07-23T16:23:42Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21447", "title": "RUMBA: Russian User Memory Benchmark", "published": "2026-07-23T15:52:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21412", "title": "Euclid-MCP: A Model Context Protocol Server for Deterministic Logical Reasoning via Prolog", "published": "2026-07-23T15:15:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21356", "title": "Emergent Misalignment Recruits a Pre-existing Persona Subspace", "published": "2026-07-23T14:19:28Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21340", "title": "Capital Markets LLM Reliability Score (CM-LRS): From Plausible to Bankable", "published": "2026-07-23T14:10:58Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21324", "title": "GRADRAG: Cross-Component Prompt Adaptation for Coordinated Multi-Agent RAG", "published": "2026-07-23T13:54:38Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21284", "title": "news-crawler-LM: A Small Long-Context Model For High-Quality News Crawling", "published": "2026-07-23T13:05:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21231", "title": "Progressive Cramming: Reliable Token Compression and What It Reveals", "published": "2026-07-23T11:46:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21677", "title": "Enhancing SLMs for Sustainable Code Optimization in Radio-Astronomy", "published": "2026-07-23T11:02:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21143", "title": "One More Turn, Less Regret: A Regret-Based Multi-Turn Benchmark for LLMs' Clarification Policies", "published": "2026-07-23T10:22:11Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21090", "title": "Training Large Language Models for Self-Explanation Faithfulness", "published": "2026-07-23T09:20:38Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "process-reward", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.21075", "title": "VibeVoice-ASR-BitNet Technical Report", "published": "2026-07-23T09:08:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21067", "title": "PrefReward: Learning User Preference Matrix for Personalized Text Generation", "published": "2026-07-23T09:00:20Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "process-reward", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21063", "title": "QuantiBias: Benchmarking Quantization-Induced Bias in LLMs", "published": "2026-07-23T08:56:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21051", "title": "Sample-Efficient Learning from Agent Experience", "published": "2026-07-23T08:34:31Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21019", "title": "HiMe: Real-Time Self-Hosted Personal Agent Platform for Health Insights with Wearable Devices", "published": "2026-07-23T08:05:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20999", "title": "Workflow-Localized Mechanism Learning: Attribution-Guided Repair and Knowledge Reuse for Structured Agent Skills", "published": "2026-07-23T07:28:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20982", "title": "GuardianAgentBench: Where Agents Fail and How to Guard Them", "published": "2026-07-23T07:05:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21670", "title": "Ordered Action Tokens for Visuomotor Policy Learning", "published": "2026-07-23T07:04:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20952", "title": "The Weight of Silence: A Causal Case for Weights Over the Scratchpad in Latent Chess Reasoning", "published": "2026-07-23T06:18:43Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.20938", "title": "Controllable and Content-Based Recommendations", "published": "2026-07-23T05:31:40Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20933", "title": "Transformer-Assisted LLM-Based Source Code Summarisation: to Enable More Secure Software Development", "published": "2026-07-23T05:27:24Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20914", "title": "Three-Pronged Spectral Control for Federated Parameter Efficient Fine Tuning", "published": "2026-07-23T04:49:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20911", "title": "Tencent WorkBuddy Bench: A Multi-Domain Coding-Agent Benchmark with Contamination-Resistant Task Construction", "published": "2026-07-23T04:34:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20873", "title": "LO-FAR: A Cost-Aware Local Filter for Sparse Feature Ranking in Industrial Ad Recommendation", "published": "2026-07-23T02:52:41Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24829", "title": "Improving Rare Medication Recommendation with Counterfactual Data Augmentation and Large Language Models", "published": "2026-07-23T02:48:23Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.20864", "title": "Position Bias is Hidden Behind Ceiling Effects: A Permutation Diagnostic for LLM Benchmarks", "published": "2026-07-23T02:45:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20863", "title": "Probabilistic Residual Learning for Online Recommendations", "published": "2026-07-23T02:39:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20862", "title": "CSPF: A Constrained Shared-Private Fusion Method for Non-Verifiable Preference Evaluation", "published": "2026-07-23T02:39:11Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20833", "title": "REFACT: Adaptive Fact Restatement for Compact and Faithful Chain-of-Thought Reasoning", "published": "2026-07-23T01:41:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20832", "title": "Beyond Heavy Log Curation: Perplexity-Based APT Detection via Unsupervised, Context-Augmented Language Models", "published": "2026-07-23T01:38:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20773", "title": "HARP: The Human--AI Research Platform", "published": "2026-07-22T22:38:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20737", "title": "Cardinality-Decomposed Loss: Matching Training Objectives to Relation Structure in Heterogeneous Recommendation Graphs", "published": "2026-07-22T21:33:02Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20722", "title": "REGARD: Regional Affective Differences in Large Language Models", "published": "2026-07-22T20:47:57Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20709", "title": "NVIDIA-labs OO Agents: Native Python Object-Oriented Agents", "published": "2026-07-22T20:25:55Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "agent-memory", "long-context", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20690", "title": "Learning to Detect UI Principle Violations via Reinforcement Learning", "published": "2026-07-22T19:42:59Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21655", "title": "Progress Reward Modeling for Robotic Learning: A Comprehensive Survey", "published": "2026-07-22T19:29:52Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20655", "title": "SalesLoop: Reinforcement Learning from Performance Feedback for Sales Lead Ranking", "published": "2026-07-22T18:26:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20652", "title": "Scaling Interpretable Transformers with Parity Bottleneck Layers", "published": "2026-07-22T18:25:16Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.21653", "title": "Molt: A Scalable PyTorch-Native Training Framework for Agentic Reinforcement Learning", "published": "2026-07-22T18:06:15Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20630", "title": "Demonstrating GenDB: Instance-Optimized and Customized Query Processing Code Generation via LLM Agents", "published": "2026-07-22T18:01:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20372", "title": "Notes to Self: Can LLMs Benefit from Experiential Abstractions?", "published": "2026-07-22T17:02:34Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20327", "title": "PyroDash: Cost-Efficient Token-Level Small-Large Language Model Collaborative Inference", "published": "2026-07-22T16:14:26Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20301", "title": "The Blessing of Dimensionality: How Near-Orthogonality in High-Dimensional Spaces Explains Temporal Portability", "published": "2026-07-22T15:48:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20274", "title": "Self-supervision drives representational convergence in medical foundation models more than clinical supervision", "published": "2026-07-22T15:25:05Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20268", "title": "PoTRE: Test-Time Reasoning inspired by Cognitive Heterogeneity", "published": "2026-07-22T15:20:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20216", "title": "Small, Free, and Effective: Orchestrating Open-Weight Small Language Models to Outperform Single LLM for Malware Analysis", "published": "2026-07-22T14:36:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20194", "title": "OLEDLM: A Unified Language Model for OLED Molecular Design", "published": "2026-07-22T14:16:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20145", "title": "SLAI T-Rex: Full-Parameter Post-training of the DeepSeek-V4 Family on Ascend SuperPOD", "published": "2026-07-22T13:49:17Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20090", "title": "Reinforcement Learning for Large Language Model Selective Evidence Adoption from Contaminated Retrieval Results", "published": "2026-07-22T12:45:04Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "reward-model" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.27231", "title": "KernelGenBench: A Multi-Source and Multi-Chip Benchmark for LLM-based Kernel Generation", "published": "2026-07-22T12:40:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20084", "title": "Non--negative matrix factorization using the \\textit{R} package \\textsf{nnmf}", "published": "2026-07-22T12:35:52Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20082", "title": "The Two-Process Theory of Machine Self-Report", "published": "2026-07-22T12:35:36Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20064", "title": "PRO-LONG: Programmatic Memory Enables Long-Horizon Reasoning", "published": "2026-07-22T12:11:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20062", "title": "Solar Open 2 Technical Report", "published": "2026-07-22T12:08:41Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20057", "title": "Antigen-specific Antibody Multi-modal Foundation Model for Functional Antibody Design", "published": "2026-07-22T11:59:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19987", "title": "UniRank: Benchmarking Ranking Models for Unified Sequential Modeling and Feature Interaction", "published": "2026-07-22T10:20:40Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19957", "title": "HijackKV: New Threat in Position-Independent KV Cache Reuse", "published": "2026-07-22T09:32:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19956", "title": "When Does Knowledge Distillation Hurt? Reliability-Aware Distillation for Low-Resource Language Summarization", "published": "2026-07-22T09:32:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19954", "title": "A Multi-Dimensional Evaluation of Explainability in Media Bias Detection", "published": "2026-07-22T09:29:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19935", "title": "MOF-Sleuth: Tool-Grounded Reward Alignment for Explainable Fine-Grained MOF CIF Auditing", "published": "2026-07-22T09:05:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19932", "title": "Efficient Chain-of-Modality Reasoning via Progressive Compression for Spoken Language Models", "published": "2026-07-22T09:04:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19913", "title": "JANUS: Foreseeing Latent Risk for Long-Horizon Agent Safety", "published": "2026-07-22T08:43:43Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "reward-model", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22729", "title": "Open Your Model's Eyes: Video and Context-Aware Multimodal Backchannel Prediction", "published": "2026-07-22T08:10:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19847", "title": "Auto-Fill: Learning to Predict Missing Values Accurately with Specialist Language Models", "published": "2026-07-22T07:32:09Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28661", "title": "Are the Financial Reasoning from LLMs Credible? A Real World Test over Long-Horizon Statements", "published": "2026-07-22T06:53:07Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19802", "title": "Zero-Observation User Reactivation with Gap-Driven Dimensional Gating", "published": "2026-07-22T06:32:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19739", "title": "Personalized Recommendation Tool Learning via Autonomous Language Agents", "published": "2026-07-22T04:10:41Z", "tracks": [ "agent", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22724", "title": "Progress-conditioned Group Policy Optimization for Long-Horizon Agentic Tasks", "published": "2026-07-22T04:01:56Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19718", "title": "Lightweight Person-Place Relation Extraction from Historical Newspapers with Dependency Graphs and Proximity Features", "published": "2026-07-22T03:35:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19712", "title": "How Fast Can Reward Models Score? A Systems Study of C++ and PyTorch Inference Runtimes for RLHF", "published": "2026-07-22T03:27:21Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19704", "title": "Efficient Clustering with Provable Guardrails for LLM Inference at Scale", "published": "2026-07-22T03:06:57Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence", "efficient-inference", "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19691", "title": "SLPO: Scaling Latent Reasoning via a Surrogate Policy", "published": "2026-07-22T02:45:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19595", "title": "Twin Agent: Context Residual Compression for Privilege Separated Agents", "published": "2026-07-21T21:47:52Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression", "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19506", "title": "Hybrid LLM-Guided Search for Quantum Reservoir Architecture Design", "published": "2026-07-21T18:43:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19345", "title": "Copy Less, Ground More: Overcoming Repetitive Copying in Long-Context Reasoning via Evidence-Aware Reinforcement Learning", "published": "2026-07-21T17:59:21Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "long-context", "reward-model" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19336", "title": "Agents in the Wild: Where Research Meets Deployment", "published": "2026-07-21T17:55:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19331", "title": "ISO: An RLVR-Native Optimization Stack", "published": "2026-07-21T17:51:36Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "on-policy-distillation", "opd", "post-training" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B09", "plan_reason": "Rubric、外部 rollout 与多奖励 RL — completed" }, { "arxiv_id": "2607.19253", "title": "Sequential Learner Modeling Using Multi-Relational Graph Convolutional Networks", "published": "2026-07-21T16:23:16Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19243", "title": "Inference-Time Steering for Cross-Lingual Factual Consistency in LLMs", "published": "2026-07-21T16:15:05Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11219", "title": "From Monolithic to Modular: Segment-level Automatic Prompt Optimization", "published": "2026-07-21T16:02:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.19226", "title": "The Price of Reasoning: Cost-Quality Tradeoffs in Reinforcement Learning for Neural Machine Translation", "published": "2026-07-21T15:57:36Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19223", "title": "AdaFlash: Adaptive Speculative Decoding via On-Policy Distilled Diffusion Drafters", "published": "2026-07-21T15:52:50Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22716", "title": "Visual Token Compression Enhances Robustness of MLLMs", "published": "2026-07-21T15:52:30Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19219", "title": "Beyond Score Prediction: LLM-Based Essay Scoring and Feedback Generation via Reinforcement Learning with Rubric Rewards", "published": "2026-07-21T15:49:02Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-rl", "post-training", "preference-optimization" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.19198", "title": "ATLAS: A Foundation Neural Sampler for Amorphous Materials", "published": "2026-07-21T15:31:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19190", "title": "Agentic Real2Sim: Physics-based World Modeling with Vision-Language Agents", "published": "2026-07-21T15:23:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19189", "title": "Spectral Biclustering-Driven Scalability for Post-Hoc Explainability in Recommender Systems", "published": "2026-07-21T15:22:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19181", "title": "Reasoning Before Translation: Enhancing Legal Machine Translation with Structured Reasoning", "published": "2026-07-21T15:15:56Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19128", "title": "One Model, Many Graphs: Learning over Attributed Graphs across Heterogeneous Modalities with Vision-Language Models", "published": "2026-07-21T14:18:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19450", "title": "REGEN: Replay-recycling for Expert-to-Generalist distillation with Offline Reinforcement Learning", "published": "2026-07-21T13:36:23Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-memory", "efficient-inference", "llm-rl", "multi-agent", "on-policy-distillation", "opd", "post-training", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.19038", "title": "FilmWorld: Agentic Novel-to-Film Generation through Dynamic Cinematic World Modeling", "published": "2026-07-21T12:28:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19031", "title": "Unsupervised Multi-kernel Learning for Automated Algorithm Selection", "published": "2026-07-21T12:18:49Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18985", "title": "Athena-Brain Technical Report: An Efficient Robot Brain for General Intelligence and Embodied Interaction", "published": "2026-07-21T11:19:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18966", "title": "Measuring Reward-Seeking via Contrastive Belief Updates", "published": "2026-07-21T10:57:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18915", "title": "Reasoning Error from Known Fact: Step-Level Self-Consistency Group Relative Policy Optimization for LLM", "published": "2026-07-21T09:56:58Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18912", "title": "From a Multilingual Streaming ASR Backbone to Kenyan-Language Systems: Data-Centric Adaptation of Nemotron 3.5 for Kikuyu, Dholuo, and Kalenjin", "published": "2026-07-21T09:52:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07527", "title": "DocAtlas: Long-Document Understanding as Mutable-State Interaction", "published": "2026-07-21T09:45:34Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-memory", "long-context", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.18867", "title": "HindsightBench: A Black-Box Behavioral Audit Protocol for Parametric Hindsight in Time-Indexed LLM Decision Tasks", "published": "2026-07-21T08:58:36Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18859", "title": "PhoenixRepair: Rethinking Repair Strategy Exploration in Software Agents", "published": "2026-07-21T08:49:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18806", "title": "AI Tour Meeting: Group Travel Planning by LLM Agents", "published": "2026-07-21T07:36:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18786", "title": "Beyond Noisy Signals: Dual-Level Denoising for Multi-modal Sequential Recommendation", "published": "2026-07-21T07:06:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18772", "title": "RF-Agent: A Practical Framework for Building Language Agents for RFIC Design", "published": "2026-07-21T06:53:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18609", "title": "Mitigating Matthew Effect: Multi-Hypergraph Boosted Multi-Interest Self-Supervised Learning for Conversational Recommendation", "published": "2026-07-21T01:07:40Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18604", "title": "Intelligent Multi-UAV Navigation in ITNTNs: A Hierarchical LLM Approach", "published": "2026-07-21T00:34:42Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18600", "title": "Topology-Aware Tokenization for Generative Recommendation", "published": "2026-07-21T00:27:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14626", "title": "LLM Safety Alignment in Low-Resource Languages: A Systematic Literature Review", "published": "2026-07-20T21:11:00Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18485", "title": "Trusted Credentials, Untrusted Behavior: Benchmarking LLM-Agent Security in High-Performance Computing", "published": "2026-07-20T20:16:02Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-planning", "multi-agent", "web-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.18481", "title": "Search-on-Graph-R1: Training Large Language Models to Search Knowledge Graphs with Reinforcement Learning", "published": "2026-07-20T19:58:32Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18476", "title": "Structured Output Collapses Answer Diversity Across 44 Language Models", "published": "2026-07-20T19:47:16Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18460", "title": "Competitive and Complementary Tools", "published": "2026-07-20T19:17:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18412", "title": "Scalable and Efficient Joint Spiking Embedding Predictive Architecture for Large-Scale Dynamic Graphs", "published": "2026-07-20T18:02:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18236", "title": "Patch Policy: Efficient Embodied Control via Dense Visual Representations", "published": "2026-07-20T17:59:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18213", "title": "SWE-Pruner Pro: The Coder LLM Already Knows What to Prune", "published": "2026-07-20T17:47:44Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18171", "title": "FlashRT: Agent Harness for Guiding Agents to Deploy Real-Time Multimodal Applications", "published": "2026-07-20T17:12:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18163", "title": "OR Else: A Differentiable Trust Region for Policy Optimization", "published": "2026-07-20T17:07:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18366", "title": "Operational Hallucination and Safety Drift in AI Agents", "published": "2026-07-20T17:01:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18147", "title": "LLMs and Agentic AI Systems for Smart Grids: A Tutorial on Architectures and Applications", "published": "2026-07-20T16:45:13Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-planning", "multi-agent", "tool-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.18110", "title": "LLM-as-a-Coach: Experiential Learning for Non-Verifiable Tasks", "published": "2026-07-20T16:08:49Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "model-compression", "on-policy-distillation", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18102", "title": "FinSAgent: Corpus-Aligned Multi-Agent RAG Framework for Evidence-Grounded SEC Filing Question Answering", "published": "2026-07-20T16:03:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18081", "title": "SelectInfer: Selective Neuron Loading and Computation for On-Device LLMs", "published": "2026-07-20T15:48:33Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18029", "title": "Natural Language Access to Domain-Specific Metadata: A Reusable Framework for LLM Query Generation", "published": "2026-07-20T14:59:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18006", "title": "MADA-RL: Multi-Agent Debate-Aware Reinforcement Learning for Parameter-Efficient Reasoning in Compact Models", "published": "2026-07-20T14:38:00Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17948", "title": "Towards Agentic Agent-based Models: Feasibility, Performance, and Statistical Model Checking", "published": "2026-07-20T13:49:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17946", "title": "A Geometric Perspective on Stabilizing Value Conflict Resolution", "published": "2026-07-20T13:46:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17883", "title": "Zero Hallucination, by Construction: Hallucination-Aware Layered Oversight for Trustworthy Enterprise AI", "published": "2026-07-20T12:34:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17828", "title": "When a Name Is Not a Name: A Benchmark Dataset and Distilled Reasoning for Culturally Entangled Bangla Homographs in Low-Resource LLMs", "published": "2026-07-20T11:17:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17780", "title": "ETAS: An Effect-Typed Language for Agent Systems", "published": "2026-07-20T10:11:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17770", "title": "Measuring Monosemanticity in Sparse Autoencoders via Latent Activation Coherence", "published": "2026-07-20T10:03:37Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "post-training", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17765", "title": "FIFA World Cup 2026 as a Contamination-Free Benchmark for LLM Forecasting Agents: Four Models, a Bookmaker, and 104 Matches", "published": "2026-07-20T10:00:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17733", "title": "MXSens: Sensitivity-Aware Mixed-Precision Quantization for Efficient LLM Inference", "published": "2026-07-20T09:23:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17715", "title": "C$^2$KV: Compressed and Composable KV Cache Reuse for Efficient LLM Inference", "published": "2026-07-20T09:09:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "efficient-inference", "long-context", "model-compression" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B07", "plan_reason": "LLM 架构、长上下文、KV cache 与评测基础设施 — completed" }, { "arxiv_id": "2607.17641", "title": "Verify, Repair, Repeat, or Stop? Robust Stopping for Noisy Verify-Repair Loops in LLM Agents", "published": "2026-07-20T07:52:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17598", "title": "Is Progressive Disclosure All You Need for Long-Context Agents?", "published": "2026-07-20T06:35:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17572", "title": "JAGG: Jacobian-Aggregated Group Gradient for Efficient GRPO Training of Diffusion Models", "published": "2026-07-20T05:30:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17568", "title": "CoCurve: Cross-Module Co-Pruning Curvature for Training-Free Structured LLM Pruning", "published": "2026-07-20T05:27:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17545", "title": "Retain or Consolidate? Budget-Dependent Operator Selection for Language Agent Memory", "published": "2026-07-20T04:43:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17538", "title": "D-NOVA: In-Storage Retrieval Accelerator via Dual-Bound 3D NAND-Optimized Similarity Search with Vector Adaptation", "published": "2026-07-20T04:31:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17528", "title": "Can AI Agents Really Complete RTL-to-GDS? Lessons from Benchmarking Tool-Interactive EDA Workflows", "published": "2026-07-20T04:08:06Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17524", "title": "Token-Level Off-Policy Learning for Faithful Generation Under Distribution Shift", "published": "2026-07-20T03:57:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17486", "title": "SALT: Salience-Aware Lexical Trie for Long-Context Compression", "published": "2026-07-20T02:20:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17461", "title": "HyCoRec: Hypergraph-Enhanced Multi-Preference Learning for Alleviating Matthew Effect in Conversational Recommendation", "published": "2026-07-20T01:22:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17437", "title": "Empirical Grounding Improves the Realism of LLM Agents Simulating Human Behavior During Disruptions", "published": "2026-07-19T23:06:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17419", "title": "Kernelized Linear Attention: Breaking the Capacity Wall with Symmetric Cones", "published": "2026-07-19T21:46:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17384", "title": "Quantifying Diversity of Thought: A Predictive Law of Weighted LLM Ensemble Lift", "published": "2026-07-19T19:01:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17331", "title": "Agentic ERP: Multi-Agent Large Language Model Architecture for Autonomous Enterprise Resource Planning", "published": "2026-07-19T16:40:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17299", "title": "WAR: Workload-Aware Rollouts for Synchronous Agentic Reinforcement Learning", "published": "2026-07-19T15:31:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17281", "title": "AIGB-R1: Self-Evolving Generative Auto-Bidding via Hierarchical Planner-Executor Optimization", "published": "2026-07-19T14:59:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17266", "title": "Debate-on-Graph: Reliable and Adaptive Reasoning of Large Language Model on Uncertain Knowledge Graph", "published": "2026-07-19T14:17:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17262", "title": "Should Missing Modalities Always Be Necessary to Repair for Multi-modal Sentiment Analysis?", "published": "2026-07-19T14:07:43Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12371", "title": "Multi-Agent Scheduling with LLM-Assisted Contract Net Negotiation for Stream Processing in Mobile Edge Computing", "published": "2026-07-19T13:06:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17201", "title": "Non-Asymptotic Best Policy Identification Guarantees in Online Reinforcement Learning", "published": "2026-07-19T11:48:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17184", "title": "Learning Sparse Representations of Multimodal Content for Enhanced Cold Item Recommendation", "published": "2026-07-19T10:45:57Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17164", "title": "Robust Assamese Speech Recognition through Controlled Fine-Tuning of Whisper Models", "published": "2026-07-19T09:54:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17147", "title": "SlotGuard: Stop Oversharing Private Local Context in LLM Agent Transcri", "published": "2026-07-19T09:13:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17139", "title": "SynH-Rank: Quality-Aware Code Search via Diverse Data Synthesis and Hierarchical Ranking Training", "published": "2026-07-19T08:52:36Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11215", "title": "Poor Man's Agentic Modeling: Simulating Large LLM-Agent Societies on a Laptop", "published": "2026-07-19T08:44:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17117", "title": "Persistent Sparse Autoencoders: Learning Feature Timescales in Language Models", "published": "2026-07-19T08:07:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17082", "title": "OTAP: Structure-Aware Optimal Transport for Evaluating Planning and Execution in Agent Trajectories", "published": "2026-07-19T05:23:17Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17063", "title": "When LLMs Over-Answer: Measuring and Mitigating Quality Issues in LLM-Based Hardware Description Language Question Answering", "published": "2026-07-19T04:12:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17050", "title": "EvoGUI: An Evolution-Aware Benchmark for GUI State-Transition Understanding", "published": "2026-07-19T03:29:08Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.17038", "title": "Reward-Driven LLM Agent Workflows: Synthesizing POMDP Routing and Self-Correction for Autonomous Decision-Making", "published": "2026-07-19T02:51:41Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "agent-planning", "multi-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.17019", "title": "Regularize or Localize: When Training-Time KV-Cache Geometry Pays Under Quantization", "published": "2026-07-19T01:13:09Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16989", "title": "Real-World Evaluation of an AI Agent Drafting Translational Impact Summaries", "published": "2026-07-18T22:34:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16973", "title": "TurboVec: A Case Study in Cost-Efficient Private Retrieval for Enterprise RAG via Codebook-Oblivious Quantization", "published": "2026-07-18T21:43:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16972", "title": "Training Continuous Chain of Thought Models: A Tale of Two Regimes", "published": "2026-07-18T21:32:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16916", "title": "Enhancing Personalized Bladder Cancer Treatment Through Reinforcement Learning: A Recurrent Patient State Transition Decision Support Framework", "published": "2026-07-18T18:16:44Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16903", "title": "A Method for Learning Value Systems in Generative AI", "published": "2026-07-18T17:37:50Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16872", "title": "Trace-Based On-Policy Distillation for Masked Diffusion Language Models", "published": "2026-07-18T16:25:17Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16808", "title": "Schema-Constrained Document-Level Event Argument Extraction with Lightweight LLM Fine-Tuning", "published": "2026-07-18T13:04:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16741", "title": "The Anatomy of a Truth Direction: Knowledge-Dependent Dimensionality, a Relational Law, and a Convergent Category Geometry in Small Language Models", "published": "2026-07-18T10:03:21Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16716", "title": "RECON: Benchmarking Agent Memory for Compositional Reasoning over Long Contexts", "published": "2026-07-18T09:11:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16660", "title": "From Adoption to Deployment: A Qualitative Study on AI Integration in Software Development Practice", "published": "2026-07-18T06:30:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16633", "title": "Beyond Fixed Depths and Widths: Optimizing Textual Decoding Tries in LLM-based Generative Recommendation", "published": "2026-07-18T04:31:02Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.24822", "title": "A Position Paper on Recommender Systems in the Era of Autonomous Agents", "published": "2026-07-18T01:18:18Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16553", "title": "Discrete Ricci Curvature on Protein Contact Graphs for Lightweight Fold Classification", "published": "2026-07-17T23:24:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16524", "title": "Feedback Attribution and Representation Geometry: Metrics for Comparing Individual and Shared Rewards in MARL", "published": "2026-07-17T21:53:07Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16451", "title": "Committed Before Reasoning: Behavioral Reproduction and Preliminary Activation-Level Evidence of Answer Pre-Commitment in an Open-Weight LLM", "published": "2026-07-17T18:49:15Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16431", "title": "RIMS: Preference Optimization via Smoothed Multi-pair Aggregation for Small-Scale LLM Retrieval-Augmented Generation", "published": "2026-07-17T18:24:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16175", "title": "Evaluating Open-Weight LLMs for Generating Structured Threat Information for Autonomous Vehicle Vulnerabilities", "published": "2026-07-17T17:55:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "multi-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16169", "title": "When Does Muon Help Agentic Reinforcement Learning?", "published": "2026-07-17T17:49:05Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16117", "title": "Rate-Utility Frontiers for Language Encodings: Comparing Tokens, Bytes, and Pixels Under Controlled Linguistic Content", "published": "2026-07-17T16:55:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16097", "title": "Understanding Reasoning from Pretraining to Post-Training", "published": "2026-07-17T16:31:58Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16066", "title": "LLM-Powered Agentic AI for 5G/6G Networks: A Tutorial and Survey on Architectures, Protocols, and Standardization", "published": "2026-07-17T15:47:23Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18310", "title": "Distribution-First Population Simulation: Collapse, Calibration, and Recall in Non-WEIRD LLM Persona Modeling", "published": "2026-07-17T15:45:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16057", "title": "Frontier AI performance across the business disciplines: a case-grounded benchmark of knowledge work and analytical reasoning", "published": "2026-07-17T15:34:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16051", "title": "Loop the Loopies!", "published": "2026-07-17T15:28:43Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16021", "title": "Candidate Attended Dialogue State Tracking Using BERT", "published": "2026-07-17T14:56:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16372", "title": "AoA: Theorem Proving Agent over Abstract Syntax Tree of Redesigned Language", "published": "2026-07-17T14:43:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15865", "title": "An MLIR-Based Compilation Method for Large Language Models", "published": "2026-07-17T11:24:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15829", "title": "Cost-efficient generative AI summarization for scalable automated essay scoring in educational assessment", "published": "2026-07-17T10:39:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15740", "title": "Debiasing Text-to-Image Evaluation via Implicit Cultural Alignment Reward Modeling", "published": "2026-07-17T08:23:03Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15736", "title": "Better Starts, Better Ends: Bootstrapped Iterative Self-Reasoning Distillation for Compressed Reasoning", "published": "2026-07-17T08:15:35Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15715", "title": "Behavioral Controllability of Agentic Models for Information Extraction: From Fixed Workflows to Reflective Agents", "published": "2026-07-17T07:51:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15707", "title": "From Skill Extraction to Multistakeholder Recommendation: A Two-Stage Framework for Bias Governance in Skills-Based Job Matching", "published": "2026-07-17T07:37:17Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18305", "title": "The Information Shadow: Measuring Structural Limits on What Language Models Can Learn", "published": "2026-07-17T06:04:07Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18304", "title": "TD-DPO: Difference-Aware Preference Optimization for Mitigating Sycophancy in Clinical Autism Intervention Dialogue", "published": "2026-07-17T05:25:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22688", "title": "Co-Harness: Co-Evolving Harnesses and Model Weights for LLM Agents", "published": "2026-07-17T02:39:57Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "efficient-inference", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15562", "title": "Hard Rules, Soft Preferences: Bridging Reasoning, Learning, and Optimization for Personalized Packing Checklist Generation", "published": "2026-07-17T02:11:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15525", "title": "Kolmogorov--Arnold Networks for Small Language Models", "published": "2026-07-17T00:22:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15509", "title": "LLM-Driven AutoML for Cross-Lingual Handwritten OCR: Closed-Loop Neural Architecture Search with GPT-5, GPT-4o, and Claude Sonnet 4", "published": "2026-07-16T23:43:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15498", "title": "VarRate: Training-Free Variable-Rate KV Cache Compression for Long-Context LLMs", "published": "2026-07-16T23:03:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15495", "title": "Verbalizable Representations Form a Global Workspace in Language Models", "published": "2026-07-16T22:54:30Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15440", "title": "Stochastic Reset Pathfinding: Path-Level Regret for Cascading Bandits over Graph Paths", "published": "2026-07-16T20:20:18Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.15380", "title": "Large Language Models as Unified Multimodal Learners for Clinical Prediction", "published": "2026-07-16T18:28:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15275", "title": "RoboTTT: Context Scaling for Robot Policies", "published": "2026-07-16T17:59:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15267", "title": "Pretraining Data Can Be Poisoned through Computational Propaganda", "published": "2026-07-16T17:56:05Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15263", "title": "Beyond Success Rate: Cost-Aware Evaluation of Offensive and Defensive Security Agents", "published": "2026-07-16T17:54:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15257", "title": "SearchOS-V1: Towards Robust Open-Domain Information-Seeking Agent Collaboration", "published": "2026-07-16T17:51:23Z", "tracks": [ "agent" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "agent-memory", "tool-agent", "web-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.15161", "title": "On-Policy Delta Distillation", "published": "2026-07-16T16:07:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15114", "title": "CoSimRec: Measuring Coordinated-Content Penetration in Recommender Feedback Loops", "published": "2026-07-16T15:24:09Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15095", "title": "Digital Pantheon: Simulating and Auditing Coalition Formation with LLM Agents", "published": "2026-07-16T15:08:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14989", "title": "OmniaBench: Benchmarking General AI Agents Across Diverse Scenarios", "published": "2026-07-16T13:38:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14967", "title": "Latent Trajectory Discrimination for AI-Generated Text Detection", "published": "2026-07-16T13:19:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14865", "title": "The Energy Society: A Simulation Environment for Studying Agent Cooperation under Survival Pressure", "published": "2026-07-16T11:40:18Z", "tracks": [ "agent", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14835", "title": "LLM-Based Re-Ranking for Real Estate Search", "published": "2026-07-16T10:58:32Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "llm-recommendation", "production-evidence", "recommendation-ranking", "recsys-general", "search-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B02", "plan_reason": "7 月工业生成推荐、Agent harness 与搜索 — completed" }, { "arxiv_id": "2607.14713", "title": "Does Multi-Agent Debate Improve AI Feedback on Research Papers?", "published": "2026-07-16T08:24:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14707", "title": "Harnessing LLMs for Reliable Academic Supervision: A Comparative Study", "published": "2026-07-16T08:14:27Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14682", "title": "Stop Thinking, Start Looking: Efficient Post-Training for Multimodal Document Question Answering via Reasoning-Free Alignment", "published": "2026-07-16T07:43:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14675", "title": "An Intelligent-Cloud Edge Multimodal Interaction System for Robots", "published": "2026-07-16T07:39:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14658", "title": "TopoAgent: A Self-Evolving Topological Agent for Multimodal Scientific Reasoning", "published": "2026-07-16T07:22:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14622", "title": "ExaGEMM: Exploration Framework for CPU-Driven ML Inference via Associative In-Register Computing for Low-Bit GEMM", "published": "2026-07-16T06:38:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14618", "title": "PolyQ: Codesigning End-to-End Quantization Framework for Scalable Edge CPU LLM Inference", "published": "2026-07-16T06:31:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14614", "title": "Beyond Entropy: Correctness-Aware Advantage Shaping via Contrastive Policy Optimization", "published": "2026-07-16T06:25:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14604", "title": "Accelerating A/B-Tests with Counterfactual Estimation: Reducing Variance through Policy Overlap", "published": "2026-07-16T06:08:52Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "generative-recommendation", "priority-org-meta", "production-evidence", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.14543", "title": "SafeRelBench: A Spatial-Relation-Aware Benchmark for Process-Level Safety in VLM-Driven Embodied Agents", "published": "2026-07-16T03:59:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14522", "title": "A Continuous-Time Reinforcement Learning Framework for Fine-Tuning Discrete Diffusion Models", "published": "2026-07-16T03:19:53Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14514", "title": "VTM-Nav: Harnessing Cross-Episode Experience for Object-Goal Navigation with Hierarchical Visual-Topological Memory", "published": "2026-07-16T03:11:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14512", "title": "RetroAgent: Harnessing LLMs to Search Over Structured Memory for Agentic Retrosynthesis Planning", "published": "2026-07-16T03:05:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14506", "title": "Non-vacuous Generalization Bounds for Reinforcement Learning with Verifiable Rewards", "published": "2026-07-16T02:42:24Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14485", "title": "Step-Level Preference Learning for Generative Agents in Social Simulations", "published": "2026-07-16T01:58:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14456", "title": "Beyond Generalist LLMs: Specialist Agentic Systems for Structured Code Workflow Execution", "published": "2026-07-16T01:06:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14418", "title": "Adaptive Ad Load Design for Sponsored Search Markets: Evidence, Theory, and Deployment", "published": "2026-07-15T23:12:26Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence", "generative-recommendation" ], "full_text_review_required": false, "plan_status": "implemented-in-current-pr", "implementation_batch": "B02", "plan_reason": "7 月工业生成推荐、Agent harness 与搜索 — completed" }, { "arxiv_id": "2607.14399", "title": "Instrument Effects in Language-Model Honesty Evaluation: An Auditable Single-System Demonstration", "published": "2026-07-15T22:33:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.15313", "title": "Position: Quantum Program Generation Must Prioritize Validity Over Probabilistic Scaling", "published": "2026-07-15T21:28:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14371", "title": "Supervised Fine-Tuning vs. In-Context Learning: An Equilibrium Analysis of LLM Personalization under Congestion", "published": "2026-07-15T21:17:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14327", "title": "PReM: Learning What to Preserve and When to Refresh for Context Compression", "published": "2026-07-15T19:46:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14277", "title": "Multi-Head Latent Control: A Unified Interface for LLM Agent Decision Making", "published": "2026-07-15T18:36:54Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "post-training", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14252", "title": "MEMORA: Embodied Action Memory from Egocentric Videos for Reasoning and Planning", "published": "2026-07-15T18:12:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14249", "title": "MIDiff: Tackling Sparsity and Imbalance in Mobile Usage Generation via Multivariate-Imaging Diffusion", "published": "2026-07-15T18:07:54Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14051", "title": "Hindcast: Replaying Prediction Markets to Evaluate LLM Forecasters", "published": "2026-07-15T17:21:43Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14187", "title": "RxBrain: Embodied Cognition Foundation Model with Joint Language-Visual Reasoning and Imagination", "published": "2026-07-15T15:45:25Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13884", "title": "Experience Memory Graph: One-Shot Error Correction for Agents", "published": "2026-07-15T14:33:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14181", "title": "Quantize with Confidence? An Empirical Study of Quantization for Code Generation", "published": "2026-07-15T14:05:33Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13854", "title": "SPyCE: Skill-Policy Co-evolution for Multimodal Agents", "published": "2026-07-15T14:01:48Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "on-policy-distillation", "test-time-rl", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20553", "title": "CMI-Mem: Toward Generalizable Long-Term Memory Management via CMI-Augmented Reinforcement Learning", "published": "2026-07-15T13:40:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13753", "title": "Post-Training Shifts Confidence: A Three-Stage Analysis of How SFT, RL, and OPD Shape CoT Calibration", "published": "2026-07-15T12:12:37Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "on-policy-distillation", "opd", "post-training" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.16324", "title": "SGMCE: Segment-Grounded Morphological Concept Explanation for Malaria Parasite Species Identification in Thick Blood Smears", "published": "2026-07-15T12:12:21Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13712", "title": "Groc-PO: Grounded Context Preference Optimization for Truthful Multimodal LLMs", "published": "2026-07-15T11:28:27Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00030", "title": "SLMs as Multi-Agent Routers: A Progressive SFT and Reinforcement Learning Approach", "published": "2026-07-15T11:11:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13686", "title": "Optimal and Efficient Contextual Combinatorial Semi-bandits with General Function Approximation", "published": "2026-07-15T10:28:05Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13608", "title": "Automatic Ordinary Differential Equations Discovery For Biological Systems Using Large Language Model Powered Agentic System", "published": "2026-07-15T08:56:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13591", "title": "Memory as a Controlled Process: Learned Adaptive Memory Management for LLM Agents", "published": "2026-07-15T08:32:40Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning", "long-context", "on-policy-distillation", "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13544", "title": "Clustering algorithms for multivariate wind farm SCADA data filtering", "published": "2026-07-15T07:45:27Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13493", "title": "Personalizing Incremental Video Search with Hybrid Text and ID Embeddings", "published": "2026-07-15T06:42:50Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14165", "title": "Towards Reliable AI-Assisted Analog Design: Template-Constrained LLM Agents for SAR ADC Generation", "published": "2026-07-15T05:19:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13430", "title": "Exploring Post-Training Alignment of Small Language Models for Biomedical Data-to-Text Generation: A Case Study of Medication Leaflet", "published": "2026-07-15T04:22:25Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13423", "title": "Temperature Scaling Is Not Enough: Calibration Gaps Under Human Label Distributions", "published": "2026-07-15T03:52:06Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13418", "title": "Can We Steer the Black-Box? Towards Controllability-Centric Evaluation of Recommender Systems with Collaborative Agents", "published": "2026-07-15T03:37:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13408", "title": "Improving Text-to-Audio Instruction Following via Fine-Grained Feedback from Audio-Aware Large Language Models", "published": "2026-07-15T03:13:06Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13395", "title": "Self-Improving is Often Sudden: Enlightenment-style Finetuning for Large-Scale Models", "published": "2026-07-15T02:43:34Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13394", "title": "GFlowRL: Scaling Distribution-Matching RL to Large Language Models", "published": "2026-07-15T02:43:16Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13389", "title": "Where Should RL Post-Training Compute Go? Model Size, Search, Learning, and Feedback", "published": "2026-07-15T02:34:35Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13332", "title": "Agora: Collective and Permissionless Internet-Scale Pretraining of Large Language Models", "published": "2026-07-14T23:32:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13328", "title": "Privacy Preserving Recommender Systems Balancing Personalization with Privacy", "published": "2026-07-14T23:21:21Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13315", "title": "Meta-Learning Preferences for Multilingual LLM Alignment", "published": "2026-07-14T22:38:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13304", "title": "Where Does the Noise Come From? A Variance-Components Decomposition of Non-Determinism in LLM Brand Answers", "published": "2026-07-14T22:12:52Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14159", "title": "MemoHarness: Agent Harnesses That Learn from Experience", "published": "2026-07-14T21:22:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00028", "title": "Width, Memory, and Delay: A Resource Accounting for the Limits of Flat Multi-Agent Systems", "published": "2026-07-14T20:35:17Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13205", "title": "Adaptive Filtering of the KV Cache: Diagnosing and Correcting Structural-Role Bias in LLM Inference", "published": "2026-07-14T18:55:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13162", "title": "What Models Express, Suppress, and Resist: Auditing Open-Weight LLMs with Persona Vectors", "published": "2026-07-14T18:10:20Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13027", "title": "PalmClaw: A Native On-Device Agent Framework for Mobile Phones", "published": "2026-07-14T17:58:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13124", "title": "ShortOPD: Recovering Pruned LLMs with Short-to-Long On-Policy Distillation", "published": "2026-07-14T17:50:50Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12963", "title": "The Illusion of Robustness: Aggregate Accuracy Hides Prediction Flips under Task-Irrelevant Context", "published": "2026-07-14T17:01:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16311", "title": "Seeing What Is Actually There: PriVE-Bench and PriVE-Tools for Counterfactual Evaluation of Agentic Visual Evidence in VLMs", "published": "2026-07-14T16:45:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12946", "title": "ViHoRec: A Quality-Controlled Vietnamese Hotel Recommendation Dataset and Cold-Start Benchmark", "published": "2026-07-14T16:28:40Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12945", "title": "RecRec: Latent Interests Recursive Reasoning for Sequential Recommendation", "published": "2026-07-14T16:28:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12893", "title": "MemOps: Benchmarking Lifecycle Memory Operations in Long-Horizon Conversations", "published": "2026-07-14T15:33:44Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "agent-memory", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12856", "title": "Verifier-Based Reinforcement Fine-Tuning of Reasoning Models for Thermal Energy Storage Control", "published": "2026-07-14T15:08:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12747", "title": "Tracing Agentic Failure from the Flow of Success", "published": "2026-07-14T13:16:14Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13115", "title": "Improving Molecular Property Prediction in Small Language Models Using Graph-based Tools", "published": "2026-07-14T13:10:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14149", "title": "Enhancing Small Language Models Reasoning through Knowledge Graph Grounding", "published": "2026-07-14T13:07:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12696", "title": "Less Experts, Faster Decoding: Cost-Aware Speculative Decoding for Mixture-of-Experts", "published": "2026-07-14T12:22:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12640", "title": "A Learning-Rate-Gated Failure of GRPO in a Small Language and Vision-Language Model Web Agent: A Controlled Null and Its Mechanism", "published": "2026-07-14T11:17:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12631", "title": "Can Induced Emotion Bias LLM Behaviors in Sequential Decision Making?", "published": "2026-07-14T11:09:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12625", "title": "KnowAct-GUIClaw: Know Deeply, Act Perfectly, Personal GUI Assistant with Self-Evolving Memory and Skill", "published": "2026-07-14T11:04:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13107", "title": "DeepCormack: Fermi surface tomography using model-based data-driven algorithms", "published": "2026-07-14T10:10:13Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12550", "title": "A JoLT for the KV Cache: Near-Lossless KV Cache Compression via Joint Tucker and JL-Residual Allocation for LLMs", "published": "2026-07-14T09:23:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13104", "title": "Self-Improvements in Modern Agentic Systems: A Survey", "published": "2026-07-14T09:12:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12480", "title": "TRACE: An Operational Reasoning Schema for Auditable Agentic Commitments", "published": "2026-07-14T08:08:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12463", "title": "Function-Aware Fill-in-the-Middle as Mid-Training for Coding Agent Foundation Models", "published": "2026-07-14T07:44:26Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "post-training", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12425", "title": "Where Reasoning Matters: Rethinking Latent Reasoning in Semantic ID-based Generative Recommendation", "published": "2026-07-14T06:55:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14621", "title": "AutoMem: A Text-Gradient Recursive Self-Improvement Framework for Automated Memory Architectures Search", "published": "2026-07-14T06:41:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12395", "title": "Ring-Zero: Scaling Zero RL to a Trillion Parameters for Emergent Reasoning", "published": "2026-07-14T06:14:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14145", "title": "ToolAnchor: Anchoring Counterfactual Context to Boost Agentic Tool-use Capability", "published": "2026-07-14T06:03:39Z", "tracks": [ "agent", "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "post-training", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12267", "title": "Track, Rank, Crack: Epistemic Working Memory Scales Multi-Hop Reasoning in Language Agents", "published": "2026-07-14T02:10:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12236", "title": "Speculate with Memory: Lossless Acceleration for LLM Agents", "published": "2026-07-14T00:36:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12233", "title": "Fin-Analyst at FinMMEval 2026 Task 3: A Live Hybrid Trading Agent with LLM Specialists and Rule-Based Signals", "published": "2026-07-14T00:27:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12216", "title": "RCWT: Measuring Task-Budget Displacement from Coordination Content in LLM Calls", "published": "2026-07-13T23:31:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12215", "title": "Fine-Tuned Multi-Agent Framework for Detecting OCEAN in Life Narratives", "published": "2026-07-13T23:26:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12077", "title": "Graph Feedback Controls Consensus and Clique Formation in Open-Weight Language-Model Populations", "published": "2026-07-13T18:55:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.12051", "title": "Agentic systems for breast cancer treatment recommendations", "published": "2026-07-13T18:09:50Z", "tracks": [ "agent", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "llm-recommendation", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.13088", "title": "Securing LLMs in the Wild: Privacy and Security Challenges at the Edge", "published": "2026-07-13T16:45:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11787", "title": "Forgetting Our Way to Shared Meaning: Effects of Forgetting on Conceptual Alignment in a Non-Partnership Coordination Game", "published": "2026-07-13T16:37:17Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26074", "title": "Reproducibility in Recommender Systems: A Survey", "published": "2026-07-13T16:30:47Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11725", "title": "Time-Lag-Aware Deep Reinforcement Learning for Flexible Job-Shop Scheduling in PPVC Module Factories", "published": "2026-07-13T15:50:00Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11722", "title": "STEP: Career-Path Recommendation via Temporal and Educational Trajectory Modeling", "published": "2026-07-13T15:48:01Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.26073", "title": "Guess Where You Go: Generative Next Point-of-Interest Recommendation in Amap", "published": "2026-07-13T15:44:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "generative-recommendation", "llm-recommendation", "priority-org-alibaba", "production-evidence", "recsys-general" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B02", "plan_reason": "7 月工业生成推荐、Agent harness 与搜索 — completed" }, { "arxiv_id": "2607.11715", "title": "JobHop v2: A Large-Scale Career Trajectory Dataset from Unstructured Resumes", "published": "2026-07-13T15:42:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11689", "title": "From World Action Models to Embodied Brains: A Roadmap for Open-World Physical Intelligence", "published": "2026-07-13T15:22:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11614", "title": "Extending LLM Context via Associative Recurrent Memory", "published": "2026-07-13T14:37:24Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11606", "title": "Globally Consistent Coloring Schemes for Language Identification", "published": "2026-07-13T14:28:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11597", "title": "Beyond Benchmarks: Exposing the Hidden Crisis in Bangla Hate Speech Detection", "published": "2026-07-13T14:19:16Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11594", "title": "MAGIC: Transition-Aware Generation of Navigable Multi-Scene Game Worlds with Large Language Models", "published": "2026-07-13T14:16:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.11210", "title": "Distribird: Literature-Informed Prior Distribution Design for Bayesian Model Calibration", "published": "2026-07-13T13:43:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11505", "title": "Proxy OPD: On-Policy Distillation with Transferable Relative Proxy Update", "published": "2026-07-13T12:56:21Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "post-training" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11503", "title": "GEIS: A Generation-Evaluation-Improvement Loop of Agent Skills for Long-Form Article Generation", "published": "2026-07-13T12:54:47Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11493", "title": "Agentic Skill Optimization over Lie Algebroids", "published": "2026-07-13T12:48:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11444", "title": "UMoE:Unlocking Every Expert in Domain-Specific Training", "published": "2026-07-13T11:52:42Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11434", "title": "Direct Image-to-Modern Vietnamese Translation of Han-Nom Manuscripts via Multimodal RLHF Preference Alignment", "published": "2026-07-13T11:40:45Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11423", "title": "ToFu: A White-Box, Token-Efficient Agent Harness for Researchers", "published": "2026-07-13T11:26:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11399", "title": "Agentic Routing: The Harness-Native Data Flywheel", "published": "2026-07-13T11:05:55Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11388", "title": "StructAgent: Harness Long-horizon Digital Agents with Unified Causal Structure", "published": "2026-07-13T10:48:25Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11374", "title": "Surprisingly Simple and Effective Multi-Domain Graph Foundation Model through Graph-to-Table Alignment", "published": "2026-07-13T10:35:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11354", "title": "User Preference Induction with LLMs for Offline Top-N Recommendation Evaluation", "published": "2026-07-13T10:16:24Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11338", "title": "AutoVSR: Automatic Visual-to-Symbolic Reasoning for Symbolic Expression Generation from Circuit Schematic", "published": "2026-07-13T09:56:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11327", "title": "PRISM Edit: One Vector for All Temporal Answers", "published": "2026-07-13T09:46:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11269", "title": "Trustworthy synthetic data for campaign decision support: strategy simulation fidelity and the PolicySynth framework", "published": "2026-07-13T08:51:39Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11211", "title": "FastTPS: An Optimized Method for LLM Token Phase for AI accelerators", "published": "2026-07-13T08:02:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11207", "title": "ProgramTab: Boosting Table Reasoning of LLMs via Programmatic Paradigm", "published": "2026-07-13T07:58:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11183", "title": "Amplitude-Only FFN Intervention for Tool-Structured LLM Inference Method: Gated Evaluation Protocol, and Cross-Model Empirical Results", "published": "2026-07-13T07:28:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "tool-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11175", "title": "The Path to Self-Evolving Clinical Systems: Scaling Medical Agents from Assistance to Autonomy", "published": "2026-07-13T07:16:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11163", "title": "Unified Gradient Projection: Language-Balanced Continual Learning for Multilingual Low-Resource ASR", "published": "2026-07-13T06:57:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11138", "title": "A Formal Hierarchical Architecture for Agentic Orchestration with Stack-Based Execution and Lazy Discovery", "published": "2026-07-13T06:15:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11131", "title": "TIGER: Text-Conditioned Visual Gated Routing with Acceptance Alignment for Multimodal Speculative Decoding", "published": "2026-07-13T06:06:27Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "model-compression", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11119", "title": "VIA: Visual Interface Agent for Robot Control", "published": "2026-07-13T05:52:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11977", "title": "Optimization Is Not All You Need", "published": "2026-07-13T05:32:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11074", "title": "ResearchQA: Benchmarking Citation-Grounded Question-Answering on Scientific Papers", "published": "2026-07-13T04:26:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11976", "title": "LiteTopK: Exploiting the Curse of Dimensionality for a Fused Indexer-TopK Kernel in Long-Context Sparse Attention", "published": "2026-07-13T03:53:56Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11052", "title": "Domain-Aware Scaling Laws Uncover Data Synergy", "published": "2026-07-13T03:31:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11012", "title": "EasyOPD: An Easy-to-use On-Policy Distillation Framework for Large Language Models", "published": "2026-07-13T02:27:50Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10915", "title": "Normative Alignment of Recommender Systems via Internal Label Shift", "published": "2026-07-12T20:43:22Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10910", "title": "ZoRRO: A Zero-Weight Personalized Recommender System for Scalable News Recommendation", "published": "2026-07-12T20:30:24Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.10891", "title": "SETA: Scaling Environments for Terminal Agents", "published": "2026-07-12T19:40:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10805", "title": "Diagnosing and Mitigating Thinking Collapse in On-Policy Self-Distillation", "published": "2026-07-12T15:24:25Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10745", "title": "The First ChineseBabyLM Challenge: training data-efficient and cognitively plausible language models for Chinese", "published": "2026-07-12T12:56:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10582", "title": "MemDecay: Region-Aware KV Cache Eviction for Efficient LLM Agent Inference", "published": "2026-07-12T05:35:26Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning", "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10559", "title": "Large language model agents accelerate inverse design of metal-organic frameworks for gas separation", "published": "2026-07-12T04:18:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10557", "title": "UNIBROWSE: A Data-to-Agent Framework for Multimodal BrowseComp", "published": "2026-07-12T04:13:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10555", "title": "Tool-Adaptive LLM Reranker", "published": "2026-07-12T04:04:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10541", "title": "RecRec: Recursive Refinement for Sequential Recommendation", "published": "2026-07-12T02:53:17Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10522", "title": "Towards Autonomous and Auditable Medical Imaging Model Development", "published": "2026-07-12T00:54:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10474", "title": "Reinforcement Learning with Verifiable Physics: Post-training LLMs with Continuous Rewards", "published": "2026-07-11T20:53:04Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10441", "title": "Context by Distinct Information: An Auditable Dirichlet-Process Working Memory for Long, Redundant Context Streams", "published": "2026-07-11T18:58:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10428", "title": "Enjoy Your Talk: A Human-Centered Benchmark for Multi-Turn Dialogue with Decoupled User Simulation, Target Modeling, and Judging", "published": "2026-07-11T18:22:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10386", "title": "Structured Thoughts For Improved Reasoning And Context Pruning", "published": "2026-07-11T16:29:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10371", "title": "GigaAM Multilingual: Foundation Model for Underrepresented Languages", "published": "2026-07-11T15:48:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11948", "title": "Ontology-Amplified Distillation and Contextuality Auditing for Sovereign Enterprise Language Models: A Combined Proof-of-Mechanism and Negative-Results Method Study", "published": "2026-07-11T15:42:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10310", "title": "PolyInterview: An LLM-based Platform for Immersive Mock Interview Practice with Comprehensive Multimodal Assessment", "published": "2026-07-11T13:33:50Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10275", "title": "Information-seeking failures of large language models in agentic clinical reasoning", "published": "2026-07-11T12:10:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11946", "title": "Hybrid Continual Learning for Low-Resource Australian Aboriginal Language Identification", "published": "2026-07-11T11:01:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10239", "title": "Multilingual Semantic Retrieval for Apple Music Search", "published": "2026-07-11T10:00:48Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2608.14617", "title": "Calibrated Trust, Not Sharper Prediction: An Empirical Test of Uncertainty Fusion", "published": "2026-07-11T09:58:09Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10235", "title": "Consensus vs. Dissent: Dynamic LLM Modeling of Subjective Preferences in Group Recommenders", "published": "2026-07-11T09:48:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.10183", "title": "Automated Tensor Scheduling for Hybrid CPU-GPU LLM Inference on Consumer Devices", "published": "2026-07-11T08:00:05Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10180", "title": "ActiveFly-Bench: Aligning Embodied Question Answering with Vision-Language-Action for Aerial Embodied Perception", "published": "2026-07-11T07:58:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10127", "title": "GAE: Graph-Augmented Evolution for Scientific Discovery via Reinforcement Optimization", "published": "2026-07-11T05:32:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11933", "title": "Transforming LLMs into Efficient Cross-Encoders via Knowledge Distillation for RAG Reranking", "published": "2026-07-11T04:31:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10114", "title": "Cost of Reasoning in non-English Languages: A Case Study on Japanese", "published": "2026-07-11T04:30:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10079", "title": "MAG: A Web-Agent Benchmark and Harness for Multimodal Action and Guide Generation", "published": "2026-07-11T02:08:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.10016", "title": "Tokenizing Numerical and Embedding Features for LLM RecSys", "published": "2026-07-10T22:36:13Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "llm-recommendation", "priority-org-amazon", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.09988", "title": "An LLM-powered Agentic Recommendation System for Connected TV Content Discovery", "published": "2026-07-10T21:28:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.09957", "title": "Workload-Driven Optimization for On-Device Real-Time Subtitle Translation", "published": "2026-07-10T20:22:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09954", "title": "A Knowledge-Based Multi-Agent Framework for Security Control Recommendation", "published": "2026-07-10T20:18:35Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09921", "title": "Global Merger-Arbitrage Forecasting with Language Models", "published": "2026-07-10T19:16:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09908", "title": "RouteRec: Strict Evaluation of Recommender-Agent Selection and Aggregation", "published": "2026-07-10T18:58:52Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09889", "title": "Remembering Distinct Items, Not Tokens: A Learnable Dirichlet-Process Cache Between State-Space Models and Attention", "published": "2026-07-10T18:28:21Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09885", "title": "Index SLM Technical Report", "published": "2026-07-10T18:19:47Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09653", "title": "VEXAIoT: Autonomous IoT Vulnerability EXploitation using AI Agents", "published": "2026-07-10T17:52:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09540", "title": "From Raw IDs to Semantic Planning: How Recommender Systems Utilize Information at Scale", "published": "2026-07-10T15:46:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09501", "title": "Normalisation-Based Likelihood Ratio Estimation for Forensic Authorship Verification", "published": "2026-07-10T15:16:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09424", "title": "A Sovereign, Open-Source Foundation Model for German and English", "published": "2026-07-10T13:51:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09415", "title": "Self-Guided Test-Time Training for Long-Context LLMs", "published": "2026-07-10T13:45:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09385", "title": "STEEL: Sparsity-Aware Fused Attention for Energy-Efficient Long-Sequence Inference on AMD's XDNA NPU", "published": "2026-07-10T13:09:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09375", "title": "Mach-Mind-4-Flash Technical Report", "published": "2026-07-10T12:57:08Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09328", "title": "WildTrace: Benchmarking Natural Evidence Trails in Long-Context Reasoning", "published": "2026-07-10T12:09:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09230", "title": "When Does Order Flow Matter? State-Dependent L2 Liquidity-State Transitions in Crypto Futures", "published": "2026-07-10T09:21:00Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09822", "title": "Memory-Conditioned Tool Calling for Camera-First Visual Agents", "published": "2026-07-10T08:40:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09195", "title": "Toward Auditable AI Scientists: A Hypothesis Evolution Protocol for LLM Agents", "published": "2026-07-10T08:39:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09175", "title": "Scoped Verification for Reliable Long-Horizon Agentic Context Evolution under Distribution Shift", "published": "2026-07-10T08:10:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09092", "title": "AgentKGV: Agentic LLM-RAG Framework with Two-Stage Training for the Fact Verification of Knowledge Graphs", "published": "2026-07-10T04:22:34Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09052", "title": "COBS: Cumulant Order Block Sparse Attention", "published": "2026-07-10T02:48:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08991", "title": "Sensitivity-Aware Thresholding and Token Routing for Activation Sparsification in Large Language Models", "published": "2026-07-09T23:40:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08971", "title": "Stochastic Linear Bandits with Partially Observed Actions", "published": "2026-07-09T22:23:17Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08940", "title": "TSRouter: Dynamic Modality-Model Selection for Time Series Reasoning", "published": "2026-07-09T21:09:05Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "efficient-inference", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08930", "title": "BlockServe: Block-Grained Continuous Batching for High-Throughput Diffusion LLM Serving", "published": "2026-07-09T20:48:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08894", "title": "GATS: Graph-Augmented Tree Search with Layered World Models for Efficient Agent Planning", "published": "2026-07-09T19:34:29Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "efficient-inference", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08768", "title": "UniClawBench: A Universal Benchmark for Proactive Agents on Real-World Tasks", "published": "2026-07-09T17:59:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08724", "title": "Latent Memory Palace: Reasoning for Control as Autoregressive Variational Inference", "published": "2026-07-09T17:30:49Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08703", "title": "MPFlow: Learning Budgeted Max-Flow Optimization on the Lightning Network with Deep Graph Reinforcement Learning", "published": "2026-07-09T17:09:20Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.08662", "title": "WebSwarm: Recursive Multi-Agent Orchestration for Deep-and-Wide Web Search", "published": "2026-07-09T16:28:49Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "long-context", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08646", "title": "UltraX: Refining Pre-Training Data at Scale with Adaptive Programmatic Editing", "published": "2026-07-09T16:18:07Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08643", "title": "BiSCo-LLM: Lookup-Free Binary Spherical Coding for Extreme Low-Bit Large Language Model Compression", "published": "2026-07-09T16:17:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08642", "title": "DominoTree: Conditional Tree-Structured Drafting with Domino for Speculative Decoding", "published": "2026-07-09T16:16:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08601", "title": "It Takes a MAESTRO To Prune Bad Experts", "published": "2026-07-09T15:32:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08497", "title": "Cognitive-structured Multimodal Agent for Multimodal Understanding, Generation, and Editing", "published": "2026-07-09T13:55:55Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning", "long-context", "web-agent" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.08409", "title": "When Synthetic Speech Is All You Have: Better Call GRPO", "published": "2026-07-09T12:34:56Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08404", "title": "DrugGen 2: A disease-aware language model for enhancing drug discovery", "published": "2026-07-09T12:29:33Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08403", "title": "Game Theory Driven Multi-Agent Framework Mitigates Language Model Hallucination", "published": "2026-07-09T12:28:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08399", "title": "Prompt Compression via Activation Aggregation", "published": "2026-07-09T12:21:44Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08395", "title": "Token-Flow Firewall: Semantic Runtime Auditing for Persistent AI Agents", "published": "2026-07-09T12:18:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08365", "title": "DaV-Gen: End-to-End Generative Retrieval via Draft-and-Verify", "published": "2026-07-09T11:22:50Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "industrial-ranking", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.08362", "title": "Echoes Across Vietnam's Highlands, Delta, and Coast: A Multilingual Corpus for Cham, Khmer, and Tay-Nung", "published": "2026-07-09T11:20:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08299", "title": "MLPTR-CC: Multi-label Pathology Test Recommendation using Classifier Chains and SHAP", "published": "2026-07-09T09:41:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08282", "title": "Multi-Agent Firewall Architecture for Privacy Protection of Sensitive Data in Interactions with Language Models", "published": "2026-07-09T09:23:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09796", "title": "Metadata-Free Meta-Reweighted Direct Preference Optimization under Noisy Preference Labels", "published": "2026-07-09T09:20:25Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08252", "title": "AutoPersonas: A Multi-Timescale Loop Engine for Open-Ended Persona Evolution", "published": "2026-07-09T08:56:30Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08202", "title": "PIT-SUN: A Deployable Empirical Marginal Transform Framework with Expectation-Consistent Recovery for Regression in Recommender Systems", "published": "2026-07-09T08:00:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08193", "title": "Open-ended Multi-agent Autocurricula via Visual Inspection of Policies with Multi-modal LLMs", "published": "2026-07-09T07:48:36Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08161", "title": "SQuaD-SQL: Efficient Text-to-SQL with Small Language Models via LLM-Guided Knowledge Distillation", "published": "2026-07-09T06:56:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08147", "title": "Prismata: Confining Cross-Site Prompt Injection in Web Agents", "published": "2026-07-09T06:37:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08057", "title": "Towards Efficient Large Language Model Serving: A Survey on System-Aware KV Cache Optimization", "published": "2026-07-09T02:11:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08032", "title": "What to Keep, What to Forget: A Rate--Distortion View of Memory Compaction in LLMs and Agents", "published": "2026-07-09T01:15:03Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08028", "title": "From Prompts to Contracts: Harness Engineering for Auditable Enterprise LLM Agents", "published": "2026-07-09T01:08:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07964", "title": "KronQ: LLM Quantization via Kronecker-Factored Hessian", "published": "2026-07-08T22:34:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07946", "title": "DeepSWE: Measuring Frontier Coding Agents on Original, Long-Horizon Engineering Tasks", "published": "2026-07-08T21:45:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07858", "title": "Agentic AI and Retrieval-Augmented Models in Straight-Through Underwriting", "published": "2026-07-08T18:43:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07847", "title": "When Does Continual Learning Require Learning", "published": "2026-07-08T18:27:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07820", "title": "DeepSearch-World: Self-Distillation for Deep Search Agents in a Verifiable Environment", "published": "2026-07-08T18:03:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07707", "title": "Co-LMLM: Continuous-Query Limited Memory Language Models", "published": "2026-07-08T17:59:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07702", "title": "From Noisy Traces to Root Causes: Structural Trajectory Analysis and Causal Extraction for Agent Optimization", "published": "2026-07-08T17:57:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07696", "title": "Breaking Database Lock-in: Agentic Regeneration of High Performance Storage Readers for Database Bypass", "published": "2026-07-08T17:55:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07690", "title": "Agon: Competitive Cross-Model RL with Implicit Rival Grading of Reasoning", "published": "2026-07-08T17:49:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07678", "title": "How Data Shapes RoPE Frequency Usage: From Positional Scale Matching to Length Generalization", "published": "2026-07-08T17:38:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07673", "title": "MedPMC: A Systematic Framework for Scaling High-Fidelity Medical Multimodal Data for Foundation Models", "published": "2026-07-08T17:26:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07646", "title": "RL Post-Training Builds Compositional Reasoning Strategies", "published": "2026-07-08T17:04:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07626", "title": "Future Confidence Distillation in Large Language Models", "published": "2026-07-08T16:43:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07763", "title": "Unlocking Temporal Generalization in Hamiltonian Video Dynamics Models", "published": "2026-07-08T15:31:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07467", "title": "SpaCellAgent: A Self-Evolving LLM-Based Multi-Agent Framework for Trajectory Analysis", "published": "2026-07-08T14:31:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09786", "title": "Length Penalties Make Chain-of-Thought Less Monitorable", "published": "2026-07-08T14:18:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07409", "title": "DeLS-Spec: Decoupled Long-Short Contexts for Parallel Speculative Drafting", "published": "2026-07-08T13:41:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07401", "title": "Heterogeneity-Adaptive Diffusion Schrodinger Bridge for PET-Guided Whole-Body MRI Translation", "published": "2026-07-08T13:35:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07388", "title": "TF-Engram: A Train-Free Engram with SSD-Backed Memory for Large Language Models", "published": "2026-07-08T13:19:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07304", "title": "Nonlinear Bandit", "published": "2026-07-08T11:47:08Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07267", "title": "Billions of Sketches Reveal Hidden Cultural Variation in Human Concepts", "published": "2026-07-08T10:51:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07748", "title": "Selective Left-Shift: Turning Test-Time Compute and Difficulty-based Curation into Training Data for Low-Resource Code Generation", "published": "2026-07-08T09:18:56Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07144", "title": "Fractal KV-Cache Archives: Lossless Symbolic Storage with In-Place Retrieval for Long-Context LLM Inference", "published": "2026-07-08T08:37:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07108", "title": "Seeing and Reflecting: Multimodal Memory-Enhanced Agent Collaboration for Recommendation", "published": "2026-07-08T07:50:03Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07103", "title": "A knowledge-augmented dataset of high-risk driving scenarios with LLM annotations for autonomous driving", "published": "2026-07-08T07:39:21Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.16897", "title": "CityReal: Human-Aligned Urban Behavior and City Dynamics Simulation with Large-Scale LLM Agents", "published": "2026-07-08T07:10:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07050", "title": "When Top-K Misses the Decision: Tool-Call Drift in Multi-Teacher On-Policy Distillation", "published": "2026-07-08T06:26:13Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07740", "title": "Jet-Long: Efficient Long-Context Extension with Dynamic Bifocal RoPE", "published": "2026-07-08T06:23:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.07026", "title": "Constrained Decoding for Diffusion Language Models via Efficient Inference over Finite Automata", "published": "2026-07-08T05:48:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06979", "title": "Robust Federated Learning Under Real-World Client Churn", "published": "2026-07-08T04:02:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.06964", "title": "End-to-End LLM Flight Planning with RAG-based Memory and Multi-modal Coach Agent", "published": "2026-07-08T03:40:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06963", "title": "Large Language Models (LLMs) and Generative AI in Cybersecurity and Privacy: A Survey of Dual-Use Risks, AI-Generated Malware, Explainability, and Defensive Strategies", "published": "2026-07-08T03:40:26Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "llm-recommendation", "priority-org-amazon", "priority-org-google", "priority-org-meta", "priority-org-microsoft" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.06875", "title": "Video2Reaction: Mapping Video to Audience Reaction Distribution in the Wild", "published": "2026-07-08T00:17:20Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06855", "title": "Geometric Self-Distillation for Reasoning Generalization", "published": "2026-07-07T23:16:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06833", "title": "Generative Diffusion Models of Stochastic Graph Signals", "published": "2026-07-07T22:02:13Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06765", "title": "When and How to Ask: Dynamic Preference Elicitation Strategies for Conversational Recommendation", "published": "2026-07-07T19:54:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06527", "title": "RSF-GLLM: Bridging the Semantic Gap in Multi-Hop Knowledge Graph QA via Recurrent Soft-Flow and Decoupled LLM Generation", "published": "2026-07-07T17:32:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09773", "title": "EvoCUA-1.5: Online Reinforcement Learning for Multi-turn Computer-Use Agents", "published": "2026-07-07T16:36:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06461", "title": "WordVoice: Explicit and Decoupled Multi-Dimensional Word-Level Control for LLM-Based TTS", "published": "2026-07-07T16:22:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06341", "title": "Harnessing Code Agents for Automatic Software Verification", "published": "2026-07-07T14:39:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09772", "title": "A Risk-Field Enhanced Closed-Loop Digital Twin Framework for Autonomous Driving Safety Validation", "published": "2026-07-07T14:33:49Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13580", "title": "Jais 2: A Family of Arabic-Centric Open Large Language Models", "published": "2026-07-07T13:54:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06269", "title": "From Application-Layer Simulation to Native Meta-Architecture: Structural Tension as an Endogenous Driver for Heterogeneous AI Evolution", "published": "2026-07-07T13:34:27Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06175", "title": "Improving LLM-Generated Process Model Quality Through Reinforcement Learning: The Role of Reward Function Design", "published": "2026-07-07T11:53:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06160", "title": "LongCrafter: Towards Diverse Long-Context Understanding via Evidence-Graph-Guided Instruction Synthesis", "published": "2026-07-07T11:35:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06157", "title": "LLM Agents for Deliberative Collaboration: A Study on Joint Decision Making Under Partial Observability", "published": "2026-07-07T11:34:10Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06155", "title": "When Does Tool Use Increase the Expressive Power of Finite-Precision Recurrent Models?", "published": "2026-07-07T11:32:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06145", "title": "Prompting Complexity: Shortest Prompts for Texts and Behaviors in LLMs", "published": "2026-07-07T11:12:44Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06140", "title": "CurateEvo: Data-Curation Evolving for Agentic Post-Training", "published": "2026-07-07T11:07:00Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06054", "title": "BlueMagpie-TTS: A Token-Efficient Tokenizer, Language Model, and TTS for Taiwanese-Accent Code-Switching Speech", "published": "2026-07-07T09:31:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06008", "title": "PolyWorkBench: Benchmarking LLM Agents for Cross-Lingual Long-Horizon Workflows", "published": "2026-07-07T08:50:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16269", "title": "From Intent to Infrastructure: LLM-Driven Agent Compilers for ISAC Networks", "published": "2026-07-07T08:48:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05969", "title": "MemDefrag: Latent Memory Defragmentation for Large Language Models", "published": "2026-07-07T08:04:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.11918", "title": "AAAI-26 Dual Submissions: Novel Challenges", "published": "2026-07-07T06:11:08Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05863", "title": "Strategic Bargaining in Multi-Buyer Markets: Reinforcement Learning from Verifiable Rewards for LLM Negotiations", "published": "2026-07-07T05:41:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05775", "title": "Beyond the Leaderboard: A Synthesis of Tool-Use, Planning, and Reasoning Failures in Large Language Model Agents", "published": "2026-07-07T03:05:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05752", "title": "When Should LLMs Search? Counterfactual Supervision for Search Routing", "published": "2026-07-07T02:23:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05734", "title": "SCOReD: Student-Aware CoT Optimization for Recommendation Distillation", "published": "2026-07-07T01:40:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06601", "title": "TriRoute: Unified Learned Routing for Joint Adaptive Attention, Experts, and KV-Cache Allocation", "published": "2026-07-07T00:12:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05690", "title": "Memory in the Loop: In-Process Retrieval as Extended Working Memory for Language Agents", "published": "2026-07-06T23:16:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13578", "title": "BCMT: Blockwise Causal Memory Transformer", "published": "2026-07-06T22:20:55Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24804", "title": "Bumblebee: Interleaved Mixed-Layer Building Blocks for Large-Scale Recommendation Systems", "published": "2026-07-06T20:31:38Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06595", "title": "When Agents Remember Too Much: Memory Poisoning Attacks on Large Language Model Agents", "published": "2026-07-06T19:30:53Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05583", "title": "ResonatorLM: Causal Resonant Field Mixing for Efficient Long-Context Language Modeling", "published": "2026-07-06T19:28:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05577", "title": "Narrative World Model: Narratology-Grounded Writer Memory for Long-Form Fiction", "published": "2026-07-06T19:23:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05541", "title": "Self-Review Reinforcement Learning (SRRL) with Cross-Episode Memory and Policy Distillation", "published": "2026-07-06T18:24:16Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05394", "title": "Weak-to-Strong Generalization via Direct On-Policy Distillation", "published": "2026-07-06T17:59:58Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05377", "title": "Cortex: A Bidirectionally Aligned Embodied Agent Framework for Long-horizon Manipulation", "published": "2026-07-06T17:55:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05369", "title": "GaP: A Graph-as-Policy Multi-Agent Self-Learning Harness For Variational Automation Tasks", "published": "2026-07-06T17:47:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05364", "title": "REDDIT: Correcting Model-Generated Timestamp Drift in ASR without Forgetting via Replay-Based Distribution Editing", "published": "2026-07-06T17:40:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05339", "title": "TREK: Distill to Explore, Reinforce to Refine", "published": "2026-07-06T17:21:16Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05196", "title": "Unified Audio Intelligence Without Regressing on Text Intelligence", "published": "2026-07-06T15:11:57Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05184", "title": "Rethinking On-Policy Self-Distillation for Thinking Models", "published": "2026-07-06T15:01:35Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "long-context", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05174", "title": "AgentGym2: Benchmarking Large Language Model Agents in De-Idealized Real-World Environments", "published": "2026-07-06T14:56:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05155", "title": "EdgeBench: Unveiling Scaling Laws of Learning from Real-World Environments", "published": "2026-07-06T14:39:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05147", "title": "DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive Generation", "published": "2026-07-06T14:28:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05132", "title": "When Agents Lie: Premeditation, Persistence, and Exploitation in Repeated Games", "published": "2026-07-06T14:17:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05095", "title": "FAST: A Holistic Framework for Optimizing Memory-I/O, Computation, and Sampling in Temporal GNN Training", "published": "2026-07-06T13:54:01Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05029", "title": "Your Agent's Memories Are Not Its Own: Forged Reasoning Attacks on LLM Agent Memory and Defenses", "published": "2026-07-06T13:10:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.06590", "title": "AI for Cultural Heritage Textiles: Fine-Tuned Latent Diffusion for Novel Ulos Motif Synthesis", "published": "2026-07-06T08:04:46Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04763", "title": "Multi-Turn On-Policy Distillation with Prefix Replay", "published": "2026-07-06T07:56:53Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04713", "title": "RSPO: Reward-Swap Policy Optimization for Multi-Turn LLM Agents", "published": "2026-07-06T06:32:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04640", "title": "Wrong Before Right: Late Rescue and Interface Failure in Aligned Language Models", "published": "2026-07-06T03:51:24Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04574", "title": "A Few Teacher Steps Go a Long Way: Cost-Efficient On-Policy Data Augmentation for Agent Post-Training", "published": "2026-07-06T01:02:30Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04562", "title": "Heaviside Continuity of Rolling Coefficients for Eliminating Epistemic Entropy in Large Language Models", "published": "2026-07-06T00:29:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04534", "title": "Mechanism-level routing failure in LLMs over Lean-verified algebraic structures", "published": "2026-07-05T22:45:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04529", "title": "Evaluation and Explainability of Unsupervised Scholarly Collaboration Recommendations", "published": "2026-07-05T22:23:50Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.05458", "title": "Learning to Control LLM Agent Harnesses with Offline Reinforcement Learning", "published": "2026-07-05T22:11:18Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04510", "title": "Transplanting, inverting, and preventing a misalignment persona: method-conditional emergent misalignment in Qwen2.5", "published": "2026-07-05T21:23:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04433", "title": "Autonomous Information Seeking: A Roadmap for Agentic Recommender Systems", "published": "2026-07-05T17:55:05Z", "tracks": [ "agent", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory", "agent-planning", "llm-recommendation", "recsys-general" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04428", "title": "dOPSD: On-Policy Self-Distillation for Diffusion Language Models", "published": "2026-07-05T17:47:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04425", "title": "UI-MOPD: Multi-Platform On-Policy Distillation for Unified GUI Agents", "published": "2026-07-05T17:37:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04391", "title": "Memory-Orchestrated Semantic System (MOSS): An Auditable Agentic Memory Architecture", "published": "2026-07-05T16:35:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-memory" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04364", "title": "RL Forgets! Towards Continual Policy Optimization", "published": "2026-07-05T15:46:11Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04356", "title": "How Many Initial Points Does Bayesian Optimization Need?", "published": "2026-07-05T15:22:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04270", "title": "LBR: Towards Mitigating Length Bias in Large Language Models for Recommendation", "published": "2026-07-05T12:31:11Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-amazon", "recsys-general" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.04244", "title": "Quantize the Target, Quantize the Drafter: Efficient Inference with Qwen3.5-4B", "published": "2026-07-05T11:44:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04206", "title": "Sangam: Efficiently Serving Diffusion LLMs with the AR Stack", "published": "2026-07-05T09:55:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04171", "title": "Teaching Tiny VLA Models Where to Look and How to Move", "published": "2026-07-05T08:34:39Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04153", "title": "Mask-based Predictive Representations for Reinforcement Learning", "published": "2026-07-05T07:42:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04081", "title": "A Unified Framework for In-Context Learning with Causal and Masked Language Models", "published": "2026-07-05T02:23:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04068", "title": "UniSGR: Unified Framework for Semantic ID Generation and Ranking", "published": "2026-07-05T01:00:58Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.04025", "title": "Patient-Conditioned Dual Hypergraph Reasoning for Auditable Traditional Chinese Medicine Prescription Support", "published": "2026-07-04T20:48:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03918", "title": "Beyond Item Order: Temporal Gap Tokenization for Generative Recommendation with Semantic IDs", "published": "2026-07-04T15:22:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03876", "title": "AdaptiveSD A Stability-Aware, Runtime-Adaptive Speculative Decoding Framework with Multi-Policy Orchestration for CPU-Constrained LLM Inference", "published": "2026-07-04T13:39:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03863", "title": "Rethinking Scientific Discovery in the Agentic Era", "published": "2026-07-04T13:09:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03788", "title": "Tensor-Train Joint Modeling for Few-Step Discrete Diffusion", "published": "2026-07-04T09:32:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19398", "title": "HyGRL: Adaptive Hybrid Graph Reasoning for Multi-Entity Questions", "published": "2026-07-04T07:07:15Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03726", "title": "SelfMem: Self-Optimizing Memory for AI Agents", "published": "2026-07-04T06:27:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20510", "title": "Telco-GAIA: Bilingual Benchmark for Agents in Telecom Domain", "published": "2026-07-04T04:34:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03624", "title": "RADIO1D: Elastic Representations for Condensed Vision Modeling", "published": "2026-07-03T22:58:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03426", "title": "Amortising Bayesian Experimental Design for Sequential Information Gathering in LLMs", "published": "2026-07-03T15:38:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03377", "title": "Spectral Signatures of Large Language Models", "published": "2026-07-03T14:30:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22667", "title": "Reinforcement Learning for Heterogeneous Sensor Selection in Maritime Surveillance", "published": "2026-07-03T14:26:26Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03362", "title": "HGenPush: A Heterogeneous Generative Recommendation Architecture for Industrial Push Notification Systems", "published": "2026-07-03T14:18:19Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "priority-org-kuaishou" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03248", "title": "Unbiased Alignment for Large Language Models with Noisy Preferences", "published": "2026-07-03T12:04:43Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03233", "title": "Agentic and Generative AI for Open-Source Intelligence and Cyber Investigations: Taxonomy, Evaluation, Challenges, and Future Directions", "published": "2026-07-03T11:42:29Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03176", "title": "Understanding electricity consumption behaviour through Inverse Reinforcement Learning", "published": "2026-07-03T10:24:30Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03158", "title": "Which Algorithm Specification Formats Help Language Models Implement Machine Learning Algorithms?", "published": "2026-07-03T09:58:38Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03089", "title": "STELLA: Efficient Sensor-to-LLM Translation for On-Device Human Activity Recognition", "published": "2026-07-03T08:21:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "efficient-inference", "model-compression", "pretraining-data" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.03065", "title": "Spectral Rewiring for Exploration, Purification, and Model Merging", "published": "2026-07-03T07:57:31Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03057", "title": "LACE-SVD: Loss-Aware SVD with Cumulative Error Correction for LLM Compression", "published": "2026-07-03T07:46:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.03050", "title": "OmniFocus: Query-Guided Modality-Balanced Token Compression for Omni-Modal Large Language Models", "published": "2026-07-03T07:41:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20507", "title": "MiniCache: Reusable Program Caching with Small Model Interfaces for Efficient LLM Inference", "published": "2026-07-03T06:53:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02966", "title": "Distill Where the Student Goes: Teacher-Regularized RL for English-Evidence Cross-Lingual RAG", "published": "2026-07-03T05:17:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19395", "title": "From Trajectories to Prefixes: Reusing Teacher Trajectories via Replayed Prefixes and Online Continuation", "published": "2026-07-03T05:07:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02959", "title": "Incentivizing Vision Language Models to Search for Long Video Question Answering", "published": "2026-07-03T05:05:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19394", "title": "Cross-Subject Semantic Decoding with Shared-Space Alignment for Generalized Neural Representation Learning", "published": "2026-07-03T03:59:16Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02881", "title": "PraMem: Practice-derived Experiential Memory for Long-horizon Behavior Prediction", "published": "2026-07-03T02:25:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02869", "title": "Reward Granularity in RLVR: Comparing Process and Outcome Reward Structures for Mathematical Reasoning in Small Language Models", "published": "2026-07-03T02:16:06Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02854", "title": "EvoOtter: Evolutionary Reproduction Test Generator", "published": "2026-07-03T01:30:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02818", "title": "Session-Level Optimization for Large-Scale Retrieval using REINFORCE with Multi-Step Off-Policy Correction", "published": "2026-07-02T23:13:36Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02718", "title": "Diagnosing Aerial-View Object Detectors with Foundational Image Generative Models", "published": "2026-07-02T19:11:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09739", "title": "Coresets Before Score Sets: Evaluation-Unsupervised Prompt Subset Selection for LLM Benchmarks", "published": "2026-07-02T18:37:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02686", "title": "ASK in the Dark: Uncertainty-Gated LLM Assistance under Partial Observability", "published": "2026-07-02T18:26:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02513", "title": "LACUNA: A Testbed for Evaluating Localization Precision for LLM Unlearning", "published": "2026-07-02T17:59:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02502", "title": "DemoPSD: Disagreement-Modulated Policy Self-Distillation", "published": "2026-07-02T17:58:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02460", "title": "Neuron-Aware Data Selection for Annotation-Free LLM Self-Distillation", "published": "2026-07-02T17:27:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02391", "title": "WattGPU: Predicting Inference Power and Latency on Unseen GPUs and LLMs", "published": "2026-07-02T16:25:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02390", "title": "DecompRL: Solving Harder Problems by Learning Modular Code Generation", "published": "2026-07-02T16:25:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02345", "title": "SkillFuzz: Fuzzing Skill Composition for Implicit Intents Discovery in Open Skill Marketplaces", "published": "2026-07-02T15:49:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02235", "title": "Challenges and Recommendations for LLMs-as-a-Judge in Multilingual Settings and Low-Resource Languages", "published": "2026-07-02T14:34:07Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02115", "title": "Planning over Matrix-Factorization MDPs for Candidate Generation", "published": "2026-07-02T12:50:45Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01972", "title": "Object Aligner: A Configurable JSON Schema Similarity Score for Graphs, Applied to LLM Prompt Optimization", "published": "2026-07-02T10:07:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01830", "title": "Many Voices, One Reward: Multi-Role Rubric Generation for LLM Judging and Reward Modeling", "published": "2026-07-02T07:50:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01823", "title": "Self-Supervised Test-Time Tuning for Packet Loss Concealment", "published": "2026-07-02T07:45:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01767", "title": "Repair the Amplifier, Not the Symptom: Stable World-Model Correction for Agent Rollouts", "published": "2026-07-02T06:31:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01763", "title": "Denser $\\neq$ Better: Limits of On-Policy Self-Distillation for Continual Post-Training", "published": "2026-07-02T06:24:30Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01690", "title": "Epistemic Goggles: A Pretrained Module that Induces an Epistemic Frame via Gradient Editing", "published": "2026-07-02T04:31:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01517", "title": "Parameter Golf: What Really Works?", "published": "2026-07-01T22:29:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01480", "title": "Procedural Memory Distillation: Online Reflection for Self-Improving Language Models", "published": "2026-07-01T21:20:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01444", "title": "On the Utility and Factual Reliability of Pruned Mixture-of-Experts Models in the Biomedical Domain", "published": "2026-07-01T20:08:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01392", "title": "Multi-Objective Exploration and Preference Optimization via Mutual Information", "published": "2026-07-01T18:50:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01387", "title": "Bi-NAS: Towards Effective and Personalized Explanation for Recommender Systems via Bi-Level Neural Architecture Search", "published": "2026-07-01T18:47:42Z", "tracks": [ "post-training", "recommendation" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01218", "title": "The State-Prediction Separation Hypothesis", "published": "2026-07-01T17:55:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01179", "title": "QuasiMoTTo: Quasi-Monte Carlo Test-Time Scaling", "published": "2026-07-01T17:10:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01170", "title": "Diffusion-GR2: Diffusion Generative Reasoning Re-ranker", "published": "2026-07-01T17:02:20Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "priority-org-amazon", "recommendation-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.01153", "title": "Adversarial Pragmatics for AI Safety Evaluation: A Diagnostic Framework and Seed Benchmark for Language-Mediated Control", "published": "2026-07-01T16:33:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.09944", "title": "Navigation Alone Is Not Enough: Evaluating Explanatory Assistive UI Agents", "published": "2026-07-01T16:24:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.13570", "title": "Think in Latent, Explain in Language: Self-Explainable Latent Reasoning", "published": "2026-07-01T16:11:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01077", "title": "Message Passing Enables Efficient Reasoning", "published": "2026-07-01T15:35:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01061", "title": "Agentic generation of verifiable rules for deterministic, self-expanding reaction classification", "published": "2026-07-01T15:24:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.00924", "title": "Graph-Native Reinforcement Learning Enables Traceable Scientific Hypothesis Generation through Conceptual Recombination", "published": "2026-07-01T13:26:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.00862", "title": "CAT: Confidence-Adaptive Thinking for Efficient Reasoning of Large Reasoning Models", "published": "2026-07-01T12:27:14Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.00019", "title": "Uncertainty-Aware Simulation-Based Inference for Operations Research with Large Language Models", "published": "2026-07-01T10:39:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.00714", "title": "Self-conditioned Flow Map Language Models via Fixed-point Flows", "published": "2026-07-01T10:02:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "model-compression" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.00642", "title": "Coachable agents for interactive gameplay", "published": "2026-07-01T08:57:25Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.00491", "title": "MindEdit-Bench: Benchmarking Object-Level Counterfactual Spatial Reasoning in VLMs from In-the-Wild Photos", "published": "2026-07-01T06:19:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.00479", "title": "Ghost in the Kernel: In-Context Learning with Efficient Transformers via Domain Generalization", "published": "2026-07-01T06:03:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02588", "title": "Homer: Understanding Long-form Videos with Hierarchical Memory and Agentic Reasoning", "published": "2026-07-01T05:53:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.00423", "title": "Selective Test-Time Debiasing for CLIP via Reward Gating", "published": "2026-07-01T04:33:26Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.00339", "title": "TRACE: State-Aware Query Processing over Temporal Evidence Graphs for Conversational Data", "published": "2026-07-01T02:28:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.00283", "title": "What's Hidden Matters: Identifying Planning-Critical Occluded Agents using Vision-Language Models", "published": "2026-07-01T00:14:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20390", "title": "Ansari: A Retrieval-Grounded Islamic AI Assistant -- Architecture, Deployment, and Lessons from 140,000 Conversations", "published": "2026-06-30T20:17:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.01272", "title": "Benchmarking Federated Learning and Knowledge Distillation for Point Cloud Classification", "published": "2026-06-30T20:12:27Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18293", "title": "One Student, Many Teachers: Multi-Task On-Policy Distillation via Soft-Prompt Privileged Context", "published": "2026-06-30T19:56:05Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.00078", "title": "Exploring Line Bundle Standard Models with Transformers", "published": "2026-06-30T18:00:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.32034", "title": "QVal: Cheaply Evaluating Dense Supervision Signals for Long-Horizon LLM Agents", "published": "2026-06-30T17:58:23Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.32032", "title": "Reinforcement Learning with Metacognitive Feedback Elicits Faithful Uncertainty Expression in LLMs", "published": "2026-06-30T17:56:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.32027", "title": "Freeform Preference Learning for Robotic Manipulation", "published": "2026-06-30T17:54:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.32002", "title": "Self-Study Reconsidered: The Hidden Fragility of Learning from Self-Generated QA", "published": "2026-06-30T17:35:14Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31984", "title": "GR2 Technical Report", "published": "2026-06-30T17:22:22Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "industrial-ranking", "llm-recommendation", "recommendation-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.31916", "title": "Theory of Mind and Persuasion Beyond Conversation: Assessing the Capacity of LLMs to Induce Belief States via Planning and Action", "published": "2026-06-30T16:22:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31748", "title": "Addressing Over-Refusal in LLMs with Competing Rewards", "published": "2026-06-30T14:38:49Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31693", "title": "ShopX: A Foundation Model for Intent-to-Item Fulfillment in Agentic Shopping", "published": "2026-06-30T14:05:28Z", "tracks": [ "agent", "recommendation" ], "review_buckets": [ "manual-review", "p2-deferred-review" ], "matched_queries": [ "agent-planning", "deployment-evidence", "generative-recommendation", "llm-recommendation", "priority-org-taobao" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.31651", "title": "FARS: A Fully Automated Research System Deployed at Scale", "published": "2026-06-30T13:30:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31551", "title": "AutoTrainess: Teaching Language Models to Improve Language Models Autonomously", "published": "2026-06-30T12:09:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31519", "title": "RaBitQCache: Rotated Binary Quantization for KVCache in Long Context LLM Inference", "published": "2026-06-30T11:32:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31413", "title": "Learning to Select, Not Relearn: Hard-Routed Mixtures of Reasoning LoRAs", "published": "2026-06-30T09:40:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31371", "title": "Calibrating the Evaluator: Does Probability Calibration Mitigate Preference Coupling in LLM Agent Feedback Loops?", "published": "2026-06-30T09:03:24Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31270", "title": "Learning from Failure: Inference-Time Self-Improvement for Computer-Use Agents", "published": "2026-06-30T07:44:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31252", "title": "Embodied CAD: Solver-Grounded LLM Agents for Parametric B-Rep Assembly Modeling", "published": "2026-06-30T07:31:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31200", "title": "Agentic RAG-VLM: Affordance-Aware Retrieval-Augmented Generation with Self-Reflective Planning for Robotic Grasping", "published": "2026-06-30T06:30:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31167", "title": "MIRTH: Mutual-Information Reasoning with Temporal Hubs for Vision-Language-Action Agents", "published": "2026-06-30T05:57:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31163", "title": "ComplianceGate: Classifier-Gated Multi-Tier LLM Routing for Inference in Regulated Industries", "published": "2026-06-30T05:49:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31145", "title": "SeKV: Resolution-Adaptive KV Cache with Hierarchical Semantic Memory for Long-Context LLM Inference", "published": "2026-06-30T05:18:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31073", "title": "MultiUAV-Plat: An LLM-Oriented Platform, Benchmark and Framework for Multi-UAV Collaborative Task Planning", "published": "2026-06-30T03:02:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.31031", "title": "GenPage: Towards End-to-End Generative Homepage Construction at Netflix", "published": "2026-06-30T02:00:05Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-netflix", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B02", "plan_reason": "7 月工业生成推荐、Agent harness 与搜索 — completed" }, { "arxiv_id": "2606.30923", "title": "Behavior Cloning is Not All You Need: The Optimality of On-Policy Distillation for Noisy Expert Feedback", "published": "2026-06-29T21:18:21Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.30887", "title": "Training Therapeutic Judges and Multi-Agent Systems for Human-Aligned Mental Health Support", "published": "2026-06-29T20:22:25Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.30852", "title": "When Does Learning to Stop Help? A Cost-Aware Study of Early Exits in Reasoning Models", "published": "2026-06-29T19:33:42Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.30639", "title": "Self-Evolving World Models for LLM Agent Planning", "published": "2026-06-29T17:58:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.30616", "title": "Scaling the Horizon, Not the Parameters: Reaching Trillion-Parameter Performance with a 35B Agent", "published": "2026-06-29T17:50:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.30445", "title": "When Does Online Imitation Learning Help in LLM Post-Training? The Role of (Non-)Realizability Beyond Horizon", "published": "2026-06-29T15:17:42Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.30345", "title": "DRIFT: Difficulty Routing Self-DIstillation with Rhythm-Gated Exploration and Success BuFfer Training", "published": "2026-06-29T14:20:47Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.30316", "title": "Toward an Energy-Optimized Operation of Data Centers Located in Wind Farms Using Reinforcement Learning", "published": "2026-06-29T13:59:33Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.30111", "title": "Automating the Design of Embodied Agent Architectures", "published": "2026-06-29T10:45:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.30093", "title": "Efficient Retrieval-Augmented Generation via Token Co-occurrence Graphs", "published": "2026-06-29T10:29:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.30059", "title": "From Failure Taxonomy to Intervention: A Diagnostic Methodology for Industry-Scale AVLM in Video and Live-Streaming Platform Moderation", "published": "2026-06-29T09:52:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29980", "title": "Exploration and Online Transfer with Behavioral Foundation Models", "published": "2026-06-29T08:55:24Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29947", "title": "Diagnosing and Mitigating Retrieval Bottlenecks in LLM-Based Cold-Start Recommendation", "published": "2026-06-29T08:23:40Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29946", "title": "POEM: Partial-Order Enhanced Real-Time Sequential Modeling for Recommendation", "published": "2026-06-29T08:23:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "generative-recommendation", "industrial-ranking", "priority-org-kuaishou" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.29914", "title": "MemDelta: Controlled Baselines and Hidden Confounds in Agent Memory Evaluation", "published": "2026-06-29T07:51:22Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29869", "title": "ARKD: Adaptive Reinforcement Learning-Guided Bidirectional KL Divergence Distillation for Text Generation", "published": "2026-06-29T07:05:56Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29863", "title": "KbSD: Knowledge Boundary aware Self-Distillation for Behavioral Calibration in Agentic Search", "published": "2026-06-29T06:56:18Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29859", "title": "Exploring Motivations for Algorithm Mention in the Domain of Natural Language Processing: A Deep Learning Approach", "published": "2026-06-29T06:53:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29793", "title": "Fund2Persona: A Framework for Building and Refining Financial Advisor Personas from Fund Disclosure Data", "published": "2026-06-29T05:14:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29776", "title": "Towards Generalizable and Evidential Nuclear Magnetic Resonance-Based Molecular Structure Elucidation via Large Language Model Agent", "published": "2026-06-29T04:38:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29773", "title": "GLIP: Graph and LLM Joint Pretraining for Graph-Level Tasks", "published": "2026-06-29T04:30:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29762", "title": "Do Recommendation Algorithms Work When Users Are LLM Agents? A Case Study on Moltbook", "published": "2026-06-29T04:16:08Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16263", "title": "Preference-based Antibody Expression Ranking: Scaling with Large-scale Weak Supervision", "published": "2026-06-29T03:34:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29713", "title": "SEVA: Self-Evolving Verification Agent with Process Reward for Fact Attribution", "published": "2026-06-29T02:37:13Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29563", "title": "Coverage-Driven KV Cache Eviction for Efficient and Improved Inference of LLM", "published": "2026-06-28T19:04:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29526", "title": "The Mirage of Optimizing Training Policies: Monotonic Inference Policies as the Real Objective for LLM Reinforcement Learning", "published": "2026-06-28T17:40:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20379", "title": "A Survey on Foundations and Frontiers of Multimodal Agentic Frameworks: Techniques and Applications", "published": "2026-06-28T17:03:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29502", "title": "UCOB: Learning to Utilize and Evolve Agentic Skills via Credit-Aware On-Policy Bidirectional Self-Distillation", "published": "2026-06-28T17:02:18Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29481", "title": "To Reason or to Fabricate: Reasoning Without Shortcuts via Hint-Anchored Pairwise Aggregation", "published": "2026-06-28T16:21:04Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24791", "title": "From Naive RAG to Deep Agentic Retrieval: An Evolving Context Engineering Pipeline for Regulatory Compliance", "published": "2026-06-28T14:11:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29341", "title": "Monosemanticity in Recommender Systems", "published": "2026-06-28T11:16:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation", "priority-org-amazon" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.29280", "title": "Deterministic Decisions for High-Stakes AI. A Zero-Egress Pipeline with the Deployability of RAG and the Accuracy of Machine Learning", "published": "2026-06-28T08:58:17Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29171", "title": "Symbolic Mechanistic Data Attribution: Tracing Training Influence to Learned Behavioral Policies", "published": "2026-06-28T03:32:06Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29116", "title": "Characterizing Large Language Model Agentic Workflows: A Study on N8n Ecosystem", "published": "2026-06-27T23:57:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29064", "title": "Fairness Attacks on Recommender Systems", "published": "2026-06-27T19:50:23Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "generative-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.29014", "title": "Customized Generative AI Agent for Transportation Engineering Practice: A Development and Continued Pre-training Guideline", "published": "2026-06-27T17:22:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28992", "title": "Fine-Tuning General-Purpose Large Language Models for Agricultural Applications:A Reproducible Framework and Evaluation Protocol Based on Qwen3-8B", "published": "2026-06-27T16:02:50Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28980", "title": "Evidence-Based Text-Conditioned 3D CT Synthesis for Ovarian Cancer", "published": "2026-06-27T15:30:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28933", "title": "FinInvest-GTCN: Explainable Graph-Temporal-Causal Modeling for Risk-Aware Investment Decision Optimization", "published": "2026-06-27T14:09:57Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review", "priority-fulltext-review" ], "matched_queries": [ "generative-recommendation", "pretraining-data", "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.28896", "title": "A Task-Driven and Quality-Assured Agent Framework for SAR Data Generation", "published": "2026-06-27T13:00:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.07508", "title": "JaleesBench: Are AI Assistants Good Spiritual Company?", "published": "2026-06-27T11:59:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28831", "title": "HARD-KV: Head-Adaptive Regularization for Decoding-time KV Compression", "published": "2026-06-27T09:36:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24789", "title": "NEXT: Reasoning-Driven Video Recommendation via a Vision-Language Model", "published": "2026-06-27T06:51:58Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "deployment-evidence", "generative-recommendation", "llm-recommendation" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B00", "plan_reason": "PR #120 / NEXT" }, { "arxiv_id": "2606.28728", "title": "Improving Large-Scale Weakly Supervised ASR by Filtering and Selection", "published": "2026-06-27T04:27:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16249", "title": "An Agentic Interface for End-to-End Probabilistic Seismic Hazard and Risk Analysis", "published": "2026-06-27T02:29:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28620", "title": "Reproducing FACTER: Fairness via Conformal Thresholding and Prompt Repair", "published": "2026-06-26T21:37:01Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28480", "title": "TUA-Bench: A Benchmark for General-Purpose Terminal-Use Agents", "published": "2026-06-26T17:59:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24788", "title": "GLIDE: Guided Layerwise Hybrid Attention for Efficient LLM Inference", "published": "2026-06-26T17:48:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28249", "title": "HPRO: Hierarchical Progressive Reward Optimization via Preference Extraction for Emotional Text-to-Speech", "published": "2026-06-26T16:35:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28182", "title": "LLawCo: Learning Laws of Cooperation for Modeling Embodied Multi-Agent Behavior", "published": "2026-06-26T15:26:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.16246", "title": "Let the Data Decide: Supervision Analysis, Capability Trade-offs, and Adaptive Objective Routing in Continued Pre-Training via Off-Policy Distillation", "published": "2026-06-26T14:19:06Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28467", "title": "An Agentic AI Pipeline for Appliance-Level Energy Anomaly Detection and LLM-Driven Recommendations", "published": "2026-06-26T14:15:51Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28059", "title": "Fast and Feasible: Permutation-based Constrained Reranking for Revenue Maximization", "published": "2026-06-26T13:04:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.27997", "title": "Benchmarking on Tasks That Matter: Dataset Selection for Preserving Model Rankings", "published": "2026-06-26T11:50:22Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27936", "title": "Agentic AI-Powered Re-Identification: An Emerging, Scalable Threat to Mobility Microdata Privacy", "published": "2026-06-26T10:27:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27930", "title": "An LLM-Powered Semantic Alignment Framework for Journal Recommendation", "published": "2026-06-26T10:22:24Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27917", "title": "Graph Dimensionality Reduction for Contextual Bandits: Structure-Specific Regret Bounds under Approximate Smoothness and Noisy Eigenspaces", "published": "2026-06-26T10:07:10Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27806", "title": "Agent vs. Parametric World Models: Hybrid Planning for Reliable Language Agents", "published": "2026-06-26T07:45:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27791", "title": "NLL-Guided Full-Attention Layer Selection for Training-Free Sliding-Window Adaptation", "published": "2026-06-26T07:20:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22649", "title": "STAIF: A Stage-wise Optimization for Complex Instruction Following", "published": "2026-06-26T06:14:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27748", "title": "Flexformer: Flexible Linear Transformer with Learnable Attention Kernel", "published": "2026-06-26T06:08:44Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27739", "title": "The Weakest Link Tells It All: Outcome-Supervised Process Reward Modeling via Learnable Credit Assignment", "published": "2026-06-26T05:38:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27705", "title": "Mitigating Position Bias in Transformers via Layer-Specific Positional Embedding Scaling", "published": "2026-06-26T04:07:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27684", "title": "Intuition-Guided Latent Reasoning for LLM-Based Recommendation", "published": "2026-06-26T03:29:48Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27679", "title": "From Signals to Transfer: A Factorised Study of Probe-Based Uncertainty Estimation in Large Language Models", "published": "2026-06-26T03:23:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27632", "title": "Yuvion LLM: An Adversarially-Aware Large Language Model for Content And AI Safety", "published": "2026-06-26T01:12:02Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27627", "title": "HybridCodec: Modeling Discrete and Continuous Representations for Efficient Speech Language Models", "published": "2026-06-26T00:53:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27608", "title": "Qwen-Image-2.0-RL Technical Report", "published": "2026-06-25T23:49:50Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27527", "title": "Large Language Model Teaches Visual Students: Cross-Modality Transfer of Fine-Grained Conceptual Knowledge", "published": "2026-06-25T20:19:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27483", "title": "Internalizing the Future: A Unified Agentic Training Paradigm for World Model Planning", "published": "2026-06-25T19:05:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27472", "title": "Supersede: Diagnosing and Training the Memory-Update Gap in LLM Agents", "published": "2026-06-25T18:50:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27377", "title": "DanceOPD: On-Policy Generative Field Distillation", "published": "2026-06-25T17:59:58Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27330", "title": "Empowering GUI Agents via Autonomous Experience Exploration and Hindsight Experience Utilization for Task Planning", "published": "2026-06-25T17:44:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27180", "title": "Automating Potential-based Reward Shaping with Vision Language Model Guidance", "published": "2026-06-25T15:45:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27058", "title": "UniFormer: Efficient and Unified Model-Centric Scaling for Industrial Recommendation", "published": "2026-06-25T14:03:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-kuaishou", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.27025", "title": "Improving General Role-Playing Agents via Psychology-Grounded Reasoning and Role-Aware Policy Optimization", "published": "2026-06-25T13:34:47Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.26917", "title": "GEOALIGN: Geometric Rollout Curation for Robust LLM Reinforcement Learning", "published": "2026-06-25T11:53:37Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.26859", "title": "AgentX: Towards Agent-Driven Self-Iteration of Industrial Recommender Systems", "published": "2026-06-25T10:42:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "deployment-evidence", "industrial-ranking", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B00", "plan_reason": "PR #120 / AgentX" }, { "arxiv_id": "2606.26797", "title": "Reasoning Quality Emerges Early: Data Curation for Reasoning Models", "published": "2026-06-25T09:32:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.26787", "title": "AIGP: An LLM-Based Framework for Long-Term Value Alignment in E-Commerce Pricing", "published": "2026-06-25T09:21:42Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "on-policy-distillation", "opd", "preference-optimization" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.26629", "title": "From Weights to Features: SAE-Guided Activation Regularization for LLM Continual Learning", "published": "2026-06-25T05:46:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.26618", "title": "Closing the Quality Gap in Low-Resource Text-to-Speech: LoRA Fine-Tuning of VoxCPM2 for Khmer and Korean", "published": "2026-06-25T05:27:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.26614", "title": "HiLSVA: Design and Evaluation of a Human-in-the-Loop Agentic System for Scientific Visualization", "published": "2026-06-25T05:19:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.26443", "title": "WatchAct: A Benchmark for Behavior-Grounded Robot Manipulation", "published": "2026-06-24T23:13:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.26399", "title": "Geometry-Aware MCTS for Extremal Problems in Combinatorial Geometry", "published": "2026-06-24T21:34:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.26387", "title": "Staying VIGILant: Mitigating Visual Laziness via Counterfactual Visual Alignment in MLLMs", "published": "2026-06-24T21:12:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.26277", "title": "From Clicks to Intent: Cross-Platform Session Embeddings with LLM-Distilled Taxonomy for Financial Services Recommendations", "published": "2026-06-24T18:18:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.26080", "title": "Neglected Free Lunch from Post-training: Progress Advantage for LLM Agents", "published": "2026-06-24T17:54:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.26050", "title": "Natural Ungrokking: Asymmetric Control of Which Rules Survive Pretraining", "published": "2026-06-24T17:27:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25800", "title": "ROAD-VLA: Robust Online Adaptation via Self-Distillation for Vision-Language-Action Models", "published": "2026-06-24T13:17:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22644", "title": "DocHRL: A Hierarchical Reinforcement Learning Framework for Cost-Optimised Document Classification", "published": "2026-06-24T13:05:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25761", "title": "Bridging Spherical Black-Box Optimizers", "published": "2026-06-24T12:35:30Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14591", "title": "6G Native AI and Channel Foundation Models", "published": "2026-06-24T10:50:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25674", "title": "BitNet Text Embeddings", "published": "2026-06-24T10:37:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.27397", "title": "SidConArena: An Environment Evaluating Agents in Open-Ended,Positive-Sum Bargaining Game", "published": "2026-06-24T10:35:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25524", "title": "Cliff Tokens: Identifying Single-Token Failure Triggers in LLM Mathematical Reasoning", "published": "2026-06-24T08:03:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20376", "title": "TH-GNN: Heterogeneous Temporal Graph Neural Networks for LLM-Agent Shilling Attack Detection", "published": "2026-06-24T08:00:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25496", "title": "Recommendation as Generation: Unifying Personalized Video Generation and Recommendation at Industrial Scale", "published": "2026-06-24T07:23:20Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.25487", "title": "How Reliable Is Your Jailbreak Judge? Calibration and Adversarial Robustness of Automated ASR Scoring", "published": "2026-06-24T07:14:17Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25442", "title": "PolicyAlign: Direct Policy-Based Safety Alignment for Large Language Models", "published": "2026-06-24T06:10:33Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25354", "title": "Efficient and Trainable Language Model Test-Time Scaling via Local Branch Routing", "published": "2026-06-24T03:42:44Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20375", "title": "GRAFT: Adaptive DLM-Based Draft Tree Construction with Target-Distilled Edge Scoring", "published": "2026-06-24T03:06:30Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25332", "title": "Decoupling Reconnaissance and Exploitation: Measuring the Capability Boundaries of LLM-Based Web Penetration Testing", "published": "2026-06-24T02:51:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25270", "title": "Inverse Reinforcement Learning for Interpretable Keystroke Biomarkers in Parkinson's Disease", "published": "2026-06-24T01:11:05Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.02542", "title": "iFLYTEK-Embodied-Omni Technical Report", "published": "2026-06-24T00:25:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25207", "title": "ASAP: Agent-System Co-Design for Wall-Clock-Centered Auto HPO Research for ML Experiments", "published": "2026-06-23T22:00:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25195", "title": "SoK: AI Secure Code Generation: Progress, Pitfalls, and Paths Forward", "published": "2026-06-23T21:39:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20374", "title": "VA-DPO: Valence-Arousal Direct Preference Optimization for Controllable Emotion Generation in Language Models", "published": "2026-06-23T19:02:20Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25039", "title": "LLM-ACES: Closed-Loop Discovery of Dynamical Systems with LLM-Guided Adaptive Search", "published": "2026-06-23T18:00:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.24790", "title": "Grad Detect: Gradient-Based Hallucination Detection in LLMs", "published": "2026-06-23T16:46:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.25000", "title": "Geo-Strat-RL: Learning Geological Event Reasoning from Verifiable Tasks", "published": "2026-06-23T16:20:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.24998", "title": "Internal Data Repetition Destroys Language Models", "published": "2026-06-23T16:02:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.24984", "title": "Learning Diachronic Representations of Ancient Greek Letterforms", "published": "2026-06-23T14:13:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.24597", "title": "Qwen-AgentWorld: Language World Models for General Agents", "published": "2026-06-23T13:53:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.24530", "title": "NatureBench: Can Coding Agents Match the Published SOTA of Nature-Family Papers?", "published": "2026-06-23T12:58:23Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.24970", "title": "Don't Go Breaking My LLM: The Impact of Pruning Attention Layers on Explanation Faithfulness and Confidence Calibration", "published": "2026-06-23T11:07:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.24957", "title": "Dustin: Draft-Augmented Sparse Verification for Efficient Long-Context Generation with Speculative Decoding", "published": "2026-06-23T08:51:20Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.24194", "title": "Dialogue to Discovery: Attribute-Aware Preference Elicitation for Conversational Product Search Assistants", "published": "2026-06-23T06:30:31Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.24143", "title": "AsyncOPD: How Stale Can On-Policy Distillation Be?", "published": "2026-06-23T04:50:49Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.24026", "title": "Can Language Model Agents be Helpful Circuit Explainers in Mechanistic Interpretability?", "published": "2026-06-23T00:04:31Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.24004", "title": "Towards Spec Learning: Inference-Time Alignment from Preference Pairs", "published": "2026-06-22T23:21:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.23964", "title": "3D Masked Autoencoders are Robust Learners of Volumetric and Multimodal Cellular Representations for Microscopy", "published": "2026-06-22T21:45:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.23911", "title": "Scaling Dense Retrieval with LLM-Annotated Training Data: Structured Mining and Progressive Curriculum for E-Commerce Sponsored Search", "published": "2026-06-22T20:19:41Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "search-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.23687", "title": "Randomized YaRN Improves Length Generalization for Long-Context Reasoning", "published": "2026-06-22T17:59:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.24937", "title": "The Hitchhiker's Guide to Agentic AI: From Foundations to Systems", "published": "2026-06-22T17:48:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.23603", "title": "MORL-A2C: Multi-Objective Reinforcement Learning Reranker for Optimizing Healthiness in MOPI-HFRS", "published": "2026-06-22T17:06:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recommendation-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.23419", "title": "GRINQH: Graded Input-based Quantization Hierarchy for Efficient LLM Generation", "published": "2026-06-22T14:42:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.23370", "title": "FlexServe: A Fast and Secure LLM Serving System for Mobile Devices with Flexible Resource Isolation", "published": "2026-06-22T14:05:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.23165", "title": "The Language Blind Spot: How Query Language and Brand Recognition Tier Shape AI-Constructed Brand Reputation Across Twelve European Languages", "published": "2026-06-22T11:05:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09706", "title": "YUKTI: From Natural-Language Situations to Robust, Verifiable Decisions An Uncertainty-Typed Proposition IR, Assumption-Robust Pareto Frontiers, and a Regret Certificate", "published": "2026-06-22T09:29:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.23057", "title": "Who Owns the AI Recommendation? A Multi-Industry Empirical Map of Brand Category Ownership Across Large Language Models", "published": "2026-06-22T09:10:54Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.23001", "title": "EnerInfer: Energy-Aware On-Device LLM Inference", "published": "2026-06-22T08:16:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22961", "title": "LLM-as-a-Judge for Reliable and Explainable Offline Evaluation in Top-K Recommendation", "published": "2026-06-22T07:42:01Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22953", "title": "Plans Don't Persist: Why Context Management Is Load Bearing for LLM Agents", "published": "2026-06-22T07:30:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22938", "title": "Provable Benefits of RLVR over SFT for Reasoning Models: Learning to Backtrack Efficiently", "published": "2026-06-22T07:16:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28385", "title": "RoboGaze: Evaluating Robot World Models via Structured Vision-Language Analysis", "published": "2026-06-22T06:45:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22873", "title": "SingGuard: A Policy-Adaptive Multimodal LLM Guardrail with Dynamic Reasoning", "published": "2026-06-22T05:37:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22803", "title": "Towards Fast Domain Adaptation and Fine-Grained User Simulation for Evaluating Conversational Recommender Systems", "published": "2026-06-22T03:26:08Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22771", "title": "Learning Moral Diversity: Modelling Individual Perspectives in Moral Classification of Texts", "published": "2026-06-22T02:19:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22741", "title": "GRADE: Graph Representation of LLM Agent Dependency and Execution", "published": "2026-06-22T01:03:21Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22722", "title": "moBERTo: A Modern Encoder for Portuguese via Continued Pretraining of ModernBERT", "published": "2026-06-21T23:44:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22606", "title": "Sub-Billion, Super-Frontier: Small Language Models Rival Zero-Shot Frontier LLMs on General and Literary Relation Extraction", "published": "2026-06-21T17:24:31Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22589", "title": "Training-free Task Classification for Multi-Task Model Merging", "published": "2026-06-21T16:51:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22485", "title": "VADAOrchestra: Neurosymbolic Orchestration of Adaptive Reasoning Workflows", "published": "2026-06-21T13:10:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22460", "title": "Music Playlist Captioning at Scale with Large Language Models", "published": "2026-06-21T12:08:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22269", "title": "Evaluating Large Language Models for Hausa and Fongbe Machine Translation: Benchmarks, Failures, and Metric Reliability", "published": "2026-06-20T23:23:54Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22189", "title": "L20-Edu-135M: An Auditable Single-GPU Study of Data-Efficient Small Language Modeling", "published": "2026-06-20T18:42:37Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22138", "title": "BioMatrix: Towards a Comprehensive Biological Foundation Model Spanning the Modality Matrix of Sequences, Structures, and Language", "published": "2026-06-20T16:38:59Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22079", "title": "Where Does the Signal Live? A Web Data Recipe for Medical Encoder Pretraining", "published": "2026-06-20T14:54:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22043", "title": "When Does a Video-Language Model Stop Watching? Reward Strength Controls the Formation and Reversal of Visual Shortcuts in Multimodal RLVR", "published": "2026-06-20T13:48:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22019", "title": "Channel Location Constrains the Auditability of Subliminal Learning", "published": "2026-06-20T12:48:31Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "on-policy-distillation", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.22013", "title": "Load Testing for Machine Learning Model Serving Systems at Scale", "published": "2026-06-20T12:34:27Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21990", "title": "Adding Robust Code-Switching Capabilities to High Performance Multilingual ASR", "published": "2026-06-20T11:02:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21940", "title": "DevoTG: Temporal Graph Neural Networks for Modeling C. elegans Developmental Connectomics", "published": "2026-06-20T08:15:42Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21906", "title": "Deeper is Not Always Better: Mitigating the Alignment Tax via Confident Layer Decoding", "published": "2026-06-20T07:03:26Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21895", "title": "Olfactory-Inspired Sparse Combinatorial Coding for Low-Resource Named Entity Recognition", "published": "2026-06-20T06:15:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21876", "title": "Protein contacts are already in the attention: a single-forward-pass alternative to the Categorical Jacobian", "published": "2026-06-20T04:35:51Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21869", "title": "The Language-Energy Divide: Measuring Energy Costs of Multilingual LLM Inference", "published": "2026-06-20T04:12:29Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21666", "title": "Hallucination as Context Drift: Synchronization Protocols for Multi-Agent LLM Systems", "published": "2026-06-19T18:17:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14126", "title": "Interpretable Language Model for Closed-Loop Type 1 Diabetes Control", "published": "2026-06-19T18:15:06Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21638", "title": "Toward Open Weight Models Without Risks: Separating Public and Private Capabilities in LLMs", "published": "2026-06-19T17:44:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21623", "title": "A DVDrive Approach for doScenes Instructed Driving Challenge", "published": "2026-06-19T17:26:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21604", "title": "Learning to Place Guards by Reinforcement: A Geo-Free Neural Policy for the Vertex-Guard Art Gallery Problem", "published": "2026-06-19T17:06:58Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21590", "title": "Radial Basis Function Networks as Projection Heads in Self-Supervised Learning", "published": "2026-06-19T16:46:54Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.21502", "title": "Towards Pedagogically Aligned LLM Tutors for Math Mistake Remediation", "published": "2026-06-19T14:53:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21458", "title": "Post-Training Speech Enhancement Language Models with Perceptual Rewards", "published": "2026-06-19T14:14:20Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21395", "title": "Atomistic Language Models Understand and Generate Materials", "published": "2026-06-19T13:01:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09693", "title": "Depth-Entropy Guided Sampling for Training-Free LLM Reasoning", "published": "2026-06-19T12:36:23Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.14587", "title": "An Agentic Framework Using Rules and LLMs for Embedding and Annotating Descriptive Document Layouts: A Plant Science Use Case", "published": "2026-06-19T11:50:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21275", "title": "A Rank-One Popularity Component in Dot-Product Recommender Scores:Population Theory and Prior-Separation Evidence", "published": "2026-06-19T09:49:50Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-alibaba" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21262", "title": "ARCO: Adaptive Rubrics with Co-Evolution for Multi-Step LLM-Based Agents", "published": "2026-06-19T09:38:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21169", "title": "Trip+: Benchmarking Agents in Personalized Interactive Travel Planning", "published": "2026-06-19T07:17:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.09692", "title": "Reference-Based Distillation Detection in LLMs", "published": "2026-06-19T06:59:19Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21144", "title": "AdaMem: Learning What to Remember for Personalized Long-Horizon LLM Agents", "published": "2026-06-19T06:35:52Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21077", "title": "OTTER: A Red-Teaming System for Toxicity-Evading Jailbreak Prompt Optimization", "published": "2026-06-19T03:55:08Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21023", "title": "Demystifying Numerical Instability in LLM Inference: Achieving Reproducible Inference for Mission-Critical Tasks with HEAL", "published": "2026-06-19T01:21:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.21013", "title": "Agentic Time Machine as an Infrastructure for Future-Event Forecasting", "published": "2026-06-19T00:55:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20961", "title": "Is Our Benchmark Enough? An Analysis of Continual Learning for MLLMs", "published": "2026-06-18T21:58:05Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20936", "title": "Comparing Transformers and Hybrid Models at the Token Level", "published": "2026-06-18T20:57:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20910", "title": "Whose Agent Are You? Multi-Layer Fingerprinting and Attribution of Autonomous Web Agents", "published": "2026-06-18T20:07:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20832", "title": "ReLaTS: a Reinforcement Learning-based method for dynamically determining the coupling Time Step in multi-scale simulations of self-gravitating systems", "published": "2026-06-18T18:19:32Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20508", "title": "What Do Safety-Aligned LLMs Learn From Mixed Compliance Demonstrations?", "published": "2026-06-18T17:25:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20487", "title": "Beyond Global Replanning: Hierarchical Recovery for Cross-Device Agent Systems", "published": "2026-06-18T17:04:17Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20164", "title": "MedRLM: Recursive Multimodal Health Intelligence for Long-Context Clinical Reasoning, Sensor-Guided Screening, Evidence-Grounded Decision Support, and Community-to-Tertiary Referral Optimization", "published": "2026-06-18T12:30:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20041", "title": "AI Economist Agent: An Agentic Framework for Model-Grounded Economic Analysis with RAG, Knowledge Graphs, and Large Language Models", "published": "2026-06-18T10:18:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20761", "title": "Integrating Large Language Model Agents with Digital Twins for Industrial Autonomous Systems", "published": "2026-06-18T09:48:44Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20014", "title": "Hierarchical Control in Multi-Agent Games: LLM-based Planning and RL Execution", "published": "2026-06-18T09:47:06Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19852", "title": "Prompt, Plan, Extract: Zero-Shot Agentic LLMs Workflows for Lung Pathology Extraction from Clinical Narratives", "published": "2026-06-18T07:00:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "agent-planning" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19830", "title": "JAMER: Project-Level Code Framework Dataset and Benchmark on Professional Game Engines", "published": "2026-06-18T06:17:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19818", "title": "Uncertainty-Aware Reward Modeling for Stable RLHF", "published": "2026-06-18T05:46:32Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19744", "title": "Beyond Uniform Forgetting: A Study of Sequential Direct Preference Optimization Across Preference Settings", "published": "2026-06-18T03:20:41Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19659", "title": "SAGE-OPD: Selective Agent-Guided Intervention for Multi-Turn On-Policy Distillation", "published": "2026-06-17T23:58:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19658", "title": "Denoising Implicit Feedback for Cold-start Recommendation", "published": "2026-06-17T23:50:45Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-kuaishou" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19625", "title": "Capability Provenance in Language Models: A Case Study in Social Reasoning", "published": "2026-06-17T22:06:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19542", "title": "Tracking Representation Dynamics in Large Language Models with Persistent Homology", "published": "2026-06-17T19:35:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20740", "title": "VeriBound: PAC-Bayesian Generalization Bounds for Process Reward Models Trained with Formal Verification Tools", "published": "2026-06-17T19:05:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19501", "title": "DeXposure-Claw: An Agentic System for DeFi Risk Supervision", "published": "2026-06-17T18:40:08Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19475", "title": "Diffusion Language Models: An Experimental Analysis", "published": "2026-06-17T18:10:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19468", "title": "Characterizing Narrative Content in Web-scale LLM Pretraining Data", "published": "2026-06-17T18:03:34Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19464", "title": "Deontic Policies for Runtime Governance of Agentic AI Systems", "published": "2026-06-17T18:02:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19341", "title": "Native Active Perception as Reasoning for Omni-Modal Understanding", "published": "2026-06-17T17:59:56Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19327", "title": "Rethinking Reward Supervision: Rubric-Conditioned Self-Distillation", "published": "2026-06-17T17:54:04Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19297", "title": "Does VLA Even Know the Basics? Measuring Commonsense and World Knowledge Retention in Vision-Language-Action Models", "published": "2026-06-17T17:20:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19253", "title": "OneCanvas: 3D Scene Understanding via Panoramic Reprojection", "published": "2026-06-17T16:29:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19218", "title": "RECOM: A Validity Discrimination Tradeoff in Automatic Metrics for Open Ended Reddit Question Answering", "published": "2026-06-17T15:55:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19170", "title": "Dango: A Strictly L1-Only Large Language Model for Studying Second Language Acquisition", "published": "2026-06-17T15:13:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19168", "title": "Beyond Safe Data: Pretraining-Stage Alignment with Regular Safety Reflection", "published": "2026-06-17T15:11:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24779", "title": "HOBA: Hierarchical On-Policy Bidding Agents for Adaptive Online Advertising", "published": "2026-06-17T15:00:34Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19120", "title": "Seeing Before Reasoning: Decoupling Perception and Reasoning for Shortcut-Resilient Multimodal On-Policy Self-Distillation", "published": "2026-06-17T14:33:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19116", "title": "Towards an Agent-First Web: Redesigning the Web for AI Agents", "published": "2026-06-17T14:31:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19108", "title": "JourneyFormer: Encoding Airbnb Guest Journey with Sequence Modeling", "published": "2026-06-17T14:22:18Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "production-evidence", "search-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B03", "plan_reason": "6–5 月序列建模、生成搜索与多场景排序 — completed" }, { "arxiv_id": "2606.19057", "title": "Quantifying and Auditing LLM Evaluation via Positive--Unlabeled Learning", "published": "2026-06-17T13:26:04Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19005", "title": "Sumi: Open Uniform Diffusion Language Model from Scratch", "published": "2026-06-17T12:32:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18961", "title": "Be Your Own Teacher: Steering Protein Language Models via Unsupervised Reward Optimization", "published": "2026-06-17T11:42:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18954", "title": "GraphPO: Graph-based Policy Optimization for Reasoning Models", "published": "2026-06-17T11:37:54Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18910", "title": "REVES: REvision and VErification--Augmented Training for Test-Time Scaling", "published": "2026-06-17T10:37:23Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18897", "title": "SAERec: Constructing Fine-grained Interpretable Intents Priors via Sparse Autoencoders for Recommendation", "published": "2026-06-17T10:17:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18844", "title": "Learning from Your Own Mistakes: Constructing Learnable Micro-Reflective Trajectories for Self-Distillation", "published": "2026-06-17T09:24:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18814", "title": "LensKit-Auto", "published": "2026-06-17T08:34:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.18812", "title": "Reinforcement Learning Foundation Models Should Already Be A Thing", "published": "2026-06-17T08:27:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18767", "title": "Output Vector Editing for Memorization Mitigation in Large Language Models", "published": "2026-06-17T07:29:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18750", "title": "Ensuring Trustworthy Online A/B Testing: Addressing Five Key Questions on CUPED", "published": "2026-06-17T06:53:03Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-bytedance", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2607.22622", "title": "Learning When to Reason for Text-to-SQL via SFT and DPO", "published": "2026-06-17T06:31:28Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20729", "title": "LLM-Guided Test-Time Discovery of Quantum-Chemical Approximation Algorithms", "published": "2026-06-17T05:58:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18703", "title": "Contextualizing Biological Language Models across Modalities via Logit-Space Contrastive Alignment", "published": "2026-06-17T05:30:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18663", "title": "RegMix-D: Dynamic Data Mixing via Proxy Training Trajectories", "published": "2026-06-17T04:02:38Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18606", "title": "Steerable Cultural Preference Optimization of Reward Models", "published": "2026-06-17T02:10:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18557", "title": "DeFAb: A Verifiable Benchmark for Defeasible Abduction in Foundation Models", "published": "2026-06-17T00:13:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18508", "title": "MCompassRAG: Topic Metadata as a Semantic Compass for Paragraph-Level Retrieval", "published": "2026-06-16T21:50:01Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18502", "title": "Towards Scalable Customization and Deployment of Multi-Agent Systems for Enterprise Applications", "published": "2026-06-16T21:30:10Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18451", "title": "A Cross-Model VLM-Judge Protocol for Single-Image 3D Mesh Quality (and Why Cheap Proxies Fall Short)", "published": "2026-06-16T20:00:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.18216", "title": "Zone of Proximal Policy Optimization: Teacher in Prompts, Not Gradients", "published": "2026-06-16T17:46:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18195", "title": "Learning from the Self-future: On-policy Self-distillation for dLLMs", "published": "2026-06-16T17:24:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18105", "title": "OmniPlan: An Adaptive Framework for Timely and Near-Optimal Network Planning Optimization", "published": "2026-06-16T16:06:51Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18062", "title": "Security and Privacy Prompts in the Wild: What Users Ask LLMs and How LLMs Respond", "published": "2026-06-16T15:37:25Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18056", "title": "ConSA: Controllable Sparsity in Hybrid Attention via Learnable Allocation", "published": "2026-06-16T15:33:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20717", "title": "MIRAGE: Stealthy Visual Prompt Injection for Vulnerability Detection in Web Agents", "published": "2026-06-16T15:31:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.18043", "title": "Uncertainty Quantification for Flow-Based Vision-Language-Action Models", "published": "2026-06-16T15:19:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17929", "title": "PreAct: Computer-Using Agents that Get Faster on Repeated Tasks", "published": "2026-06-16T13:40:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17871", "title": "StepGuard: Guarding Web Navigation via Single-Step Calibration", "published": "2026-06-16T12:42:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17846", "title": "Qwen-RobotManip Technical Report: Alignment Unlocks Scale for Robotic Manipulation Foundation Models", "published": "2026-06-16T12:14:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17803", "title": "Continual Self-Improvement with Lightweight Experiential Latent Memories", "published": "2026-06-16T11:27:28Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17707", "title": "Do Generative Recommenders Deepen the Information Cocoon? A Closed-Loop Simulation with LLM-powered User Simulators", "published": "2026-06-16T09:17:46Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17645", "title": "Beyond Domains: Reusing Web Skills via Transferable Interaction Patterns", "published": "2026-06-16T08:04:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20498", "title": "AISE-Bench: A Full-Cycle Curated Benchmark for Information Seeking on Academic Knowledge Graphs", "published": "2026-06-16T07:59:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17524", "title": "Learning to Refine Hidden States for Reliable LLM Reasoning", "published": "2026-06-16T05:03:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17489", "title": "Online LLM Selection via Constrained Bandits with Time-Varying Demand", "published": "2026-06-16T03:58:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17443", "title": "Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems", "published": "2026-06-16T02:54:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17289", "title": "Nothing from Something: Can a Language Model Discover 0?", "published": "2026-06-15T20:54:04Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17276", "title": "On the Memorization Behavior of LLMs in Generative Recommendation: Observations, Implications, and Training Strategies", "published": "2026-06-15T20:34:49Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17199", "title": "PowerOPD: Stabilizing On-Policy Distillation with Bounded Power Transformation", "published": "2026-06-15T18:37:51Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17056", "title": "The Value Axis: Language Models Encode Whether They're on the Right Track", "published": "2026-06-15T17:59:58Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17029", "title": "DEEPRUBRIC: Evidence-Tree Rubric Supervision for Efficient Reinforcement Learning of Deep Research Agents", "published": "2026-06-15T17:52:27Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "efficient-inference", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17016", "title": "TokenPilot: Cache-Efficient Context Management for LLM Agents", "published": "2026-06-15T17:46:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16934", "title": "Exploring Extrinsic and Intrinsic Properties for Effective Reasoning with Code Interpreter", "published": "2026-06-15T16:34:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16899", "title": "Fantastic Pretraining Optimizers and Where to Find Them II: Hyperball Optimization", "published": "2026-06-15T16:09:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16838", "title": "OneRank: Unified Transformer-Native Ranking Architecture for Multi-Task Recommendation", "published": "2026-06-15T15:16:17Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.16825", "title": "Tying the Loop -- Tied Expert Layers in Mixture-of-Experts Language Models", "published": "2026-06-15T15:08:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16821", "title": "How Much Can We Trust LLM Search Agents? Measuring Endorsement Vulnerability to Web Content Manipulation", "published": "2026-06-15T15:05:25Z", "tracks": [ "agent", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28369", "title": "Multimodal and Multiscale Spatial-Temporal Semantic Search and Recommendation with AI Foundation Models", "published": "2026-06-15T15:02:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16802", "title": "LabOSBench: Benchmarking Computer Use Agents for Scientific Instrument Control", "published": "2026-06-15T14:42:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16748", "title": "MyPCBench: A Benchmark for Personally Intelligent Computer-Use Agents", "published": "2026-06-15T14:08:09Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16703", "title": "Harmonizing Semantic and Collaborative in LLMs: Reasoning-based Embedding Generator for Sequential Recommendation", "published": "2026-06-15T13:37:52Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16682", "title": "Multimodal Evaluator Preference Collapse: Cross-Modal Coupling in Self-Evolving Agents", "published": "2026-06-15T13:18:20Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16659", "title": "FraudSMSWalker: Benchmarking Agentic Large Language Models for SMS-to-Webpage Fraud Detection", "published": "2026-06-15T12:53:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16641", "title": "PIANO: Personalized Reranking via Information Aggregation Node for Music Search Optimization", "published": "2026-06-15T12:29:09Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "search-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.16602", "title": "PhysGuard: Fisher-Guided Gradient Projection for Sim-to-Real Neural PDE Surrogates", "published": "2026-06-15T11:50:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16576", "title": "Can LLM Agents Infer World Models? Evidence from Agentic Automata Learning", "published": "2026-06-15T11:23:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16497", "title": "daVinci-kernel: Co-Evolving Skill Selection, Summarization, and Utilization via RL for GPU Kernel Optimization", "published": "2026-06-15T09:58:21Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16496", "title": "REFLEX: Reflective Evolution from LLM Experience", "published": "2026-06-15T09:58:13Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16456", "title": "SPRI: SVD-Partitioned Residual Initialization for Data-Constrained MoE Upcycling", "published": "2026-06-15T09:28:24Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.17113", "title": "The Critical Role of Model Selection in Causal Inference: A Comparative Analysis of Classification Models within the InferBERT Framework for Pharmacovigilance", "published": "2026-06-15T09:01:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19371", "title": "Mitigating Scaffolding Collapse in Socratic Tutors via Representation Alignment", "published": "2026-06-15T07:55:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16352", "title": "Communication-Efficient Verifiable Attention for LLM Inference", "published": "2026-06-15T07:50:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16344", "title": "Whose hotel does the AI recommend? An algorithm audit of reputation signals in LLM-assisted hotel selection", "published": "2026-06-15T07:47:31Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16262", "title": "UXBench: Measuring the Actionability of LLM-Generated UX Critiques", "published": "2026-06-15T06:08:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16246", "title": "Demystifying Training-Time Augmentation for Data-Constrained Language Model Pretraining", "published": "2026-06-15T05:48:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16222", "title": "Latent Thought Flow: Efficient Latent Reasoning in Large Language Models", "published": "2026-06-15T05:02:46Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16219", "title": "Graphical conditional generative modeling for digital twin modeling", "published": "2026-06-15T04:56:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16215", "title": "PACT: Privileged Trace Co-Training for Multi-Turn Tool-Use Agents", "published": "2026-06-15T04:46:23Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16140", "title": "VibeThinker-3B: Exploring the Frontier of Verifiable Reasoning in Small Language Models", "published": "2026-06-15T02:57:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16100", "title": "Your \"Pro\" LLM Subscription May Actually Be \"Free\": Exposing Fingerprint Spoofing Risks in LLM Inference Services", "published": "2026-06-15T01:30:42Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "efficient-inference", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16074", "title": "PVminerLLM2: Improving Structured Extraction of Patient Voice via Preference Optimization", "published": "2026-06-15T00:18:47Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.16059", "title": "Mojo: A Promising Tool for Scalable Financial AI Efficiency", "published": "2026-06-14T23:18:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15866", "title": "STRIDE: Strategic Trajectory Reasoning via Discriminative Estimation for Verifiable Reinforcement Learning", "published": "2026-06-14T15:37:26Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15838", "title": "Intelligent Multimodal Retrieval and Reasoning for Geospatial Knowledge Discovery on the I-GUIDE Platform", "published": "2026-06-14T14:39:29Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.20345", "title": "When Vocabulary Comprehension Fails Clinical Reasoning: Evaluating Therapy Bots' Safety Risks for Generation Alpha", "published": "2026-06-14T06:11:52Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15453", "title": "A Spatio-Temporal Expert Prefetching Framework for Efficient MoE-based LLM Inference", "published": "2026-06-13T20:09:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15396", "title": "CHILLGuard: Towards Fine-Grained Chinese LLM Safety Guardrail with Scalable Data Construction and Model-aware Preference Alignment", "published": "2026-06-13T16:57:51Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15333", "title": "Replay What Matters: Off-Policy Replay for Efficient LLM Reinforcement Unlearning", "published": "2026-06-13T14:52:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15307", "title": "Adapting Reinforcement Learning with Chain-of-Thought Supervision for Explainable Detection of Hateful and Propagandistic Memes", "published": "2026-06-13T13:51:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.08786", "title": "Accelerating GPU Inference of Large Language Models with Moderately Unstructured Sparse Weight Matrices", "published": "2026-06-13T13:38:27Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15277", "title": "Guiding Federated Graph Recommendation with LLM-encoded knowledge", "published": "2026-06-13T12:30:57Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15231", "title": "Visual-Seeker: Towards Visual-Native Multimodal Agentic Search via Active Visual Reasoning", "published": "2026-06-13T10:07:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15225", "title": "Edu-Theater: A Data-Efficient Agent Framework for Scalable Learner Behavior Simulation through Staging Roll-Call", "published": "2026-06-13T09:48:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15197", "title": "StarOR: Synergizing Tree Search and Test-Time Reinforcement Learning for Optimization Modeling", "published": "2026-06-13T08:46:13Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15134", "title": "Beyond Scalar Distances: Semantic Attribute Gradients from Frozen MLLMs for Visual Embeddings", "published": "2026-06-13T05:50:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15099", "title": "Think Less, Act Early: Reinforced Latent Reasoning with Early Exit in Vision-Language-Action Models", "published": "2026-06-13T04:16:18Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15077", "title": "Risk-Aware LLM Agents for Geospatial Data Retrieval: Design and Preliminary Adversarial Evaluation", "published": "2026-06-13T03:15:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15044", "title": "Equity with Efficiency: An Empirical Study of Tokenizers for Multilingual Large Language Models", "published": "2026-06-13T01:10:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15017", "title": "Are Online Skill and Memory Modules Always Worth Their Tokens? A Budget-Constrained Study of Web Agents", "published": "2026-06-12T23:30:14Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.15007", "title": "Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning", "published": "2026-06-12T22:56:12Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14970", "title": "Zero-order Parameter-free Optimization for LMO-based Methods: Novel Approach for Efficient Fine-tuning", "published": "2026-06-12T21:46:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14932", "title": "Retrieval-as-a-Service:A System-Oriented Analysis of Industrial Retrieval Pipelines in Web Systems", "published": "2026-06-12T20:10:02Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "deployment-evidence", "industrial-ranking", "llm-recommendation" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.14694", "title": "AdaSR: Adaptive Streaming Reasoning with Hierarchical Relative Policy Optimization", "published": "2026-06-12T17:56:25Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14691", "title": "CORA: Analyzing and bridging thinking-answer gap in Multimodal RLVR via Consistency-Oriented Reasoning Alignment", "published": "2026-06-12T17:54:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14654", "title": "Abstracting Cross-Domain Action Sequences into Interpretable Workflows", "published": "2026-06-12T17:19:15Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14517", "title": "From Shield to Target: Denial-of-Service Attacks on LLM-Based Agent Guardrails", "published": "2026-06-12T14:49:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14474", "title": "Verifiable User Simulation for Search and Recommendation Systems", "published": "2026-06-12T14:09:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14460", "title": "A Computational Audit of Demographic Association Encoding in ClinicalBERT Language Predictions", "published": "2026-06-12T13:51:25Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14368", "title": "Be My Tutor: On-Policy Co-Distillation for Mutual LLM Improvement via Peer Feedback", "published": "2026-06-12T11:55:51Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14817", "title": "Combining Retrieval-Augmented Text Generation with LLMs for Reading Content Recommendations", "published": "2026-06-12T11:07:54Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "llm-recommendation", "priority-org-google", "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.14299", "title": "What Drives Test-Time Adaptation for CLIP? A Controlled Empirical Study from an Update Perspective", "published": "2026-06-12T09:35:28Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14243", "title": "Decoupled Mixture-of-Experts for Parametric Knowledge Injection", "published": "2026-06-12T08:21:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14209", "title": "Detecting undisclosed LLM-generated content in parliamentary texts", "published": "2026-06-12T07:46:50Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14179", "title": "CacheRL:Multi-Turn Tool-Calling Agents via Cached Rollouts and Hybrid Reward", "published": "2026-06-12T07:01:50Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14150", "title": "Small LLMs: Pruning vs. Training from Scratch", "published": "2026-06-12T06:24:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14142", "title": "Implicit Reasoning for Large Language Model-based Generative Recommendation", "published": "2026-06-12T06:04:05Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14127", "title": "CoRe: A Continuously Reward-Finetuned LLM Query Rewriter for Multi-Stage Context-Aware Relevance in Web-Scale Video Search", "published": "2026-06-12T05:19:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19206", "title": "Hallucination as a Feature, not a Defect: Evaluating a multi-agent architecture to transform speculative language-model outputs into testable scientific hypotheses", "published": "2026-06-11T21:33:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28360", "title": "Carolina Guide: A Multi-Agent RAG System with Institutional Guardrails for Academic Policy Assistance", "published": "2026-06-11T20:56:10Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13832", "title": "Safety-Contract Graph Multi-Agent Reinforcement Learning for Autonomous Network Security Response", "published": "2026-06-11T19:02:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.14801", "title": "QPILOTS: Efficient Test-Time Q-Steering for Flow Policies", "published": "2026-06-11T18:22:03Z", "tracks": [ "post-training" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "on-policy-distillation", "opd", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.13808", "title": "The Culture Funnel: You Can't Align What isn't in the Data", "published": "2026-06-11T18:21:10Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13795", "title": "DiPOD: Diffusion Policy Optimization without Drifting Apart", "published": "2026-06-11T18:06:04Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13680", "title": "Learning to Reason by Analogy via Retrieval-Augmented Reinforcement Fine-Tuning", "published": "2026-06-11T17:59:52Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13657", "title": "Dense Supervision, Sparse Updates: On the Sparsity and Geometry of On-Policy Distillation", "published": "2026-06-11T17:54:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13624", "title": "Beyond Uniform Tokens: Adaptive Compression for Time Series Language Models", "published": "2026-06-11T17:39:26Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13610", "title": "One Polluted Page Is Enough: Evaluating Web Content Pollution in Generative Recommenders", "published": "2026-06-11T17:24:14Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13598", "title": "Reward Modeling for Multi-Agent Orchestration", "published": "2026-06-11T17:16:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13578", "title": "LabVLA: Grounding Vision-Language-Action Models in Scientific Laboratories", "published": "2026-06-11T17:03:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13426", "title": "Accelerating Speculative Diffusions via Block Verification", "published": "2026-06-11T14:54:13Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13380", "title": "An LLM System for Autonomous Variational Quantum Circuit Design", "published": "2026-06-11T14:08:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13740", "title": "Efficient On-Device Diffusion LLM Inference with Mobile NPU", "published": "2026-06-11T12:44:57Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13227", "title": "PolyAlign: Conditional Human-Distribution Alignment", "published": "2026-06-11T11:41:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13120", "title": "EvoBrowseComp: Benchmarking Search Agents on Evolving Knowledge", "published": "2026-06-11T09:48:32Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.13001", "title": "CFALR: Collaborative Filtering-Augmented Large Language Model for Personalized Fashion Outfit Recommendation", "published": "2026-06-11T07:38:15Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12994", "title": "DeepJEB++: Foundation Model-Driven Large-Scale 3D Engineering Dataset via 2D Latent Space Augmentation", "published": "2026-06-11T07:30:56Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12984", "title": "SkillChain: Closing the Loop on Skill Evolution for Image-Based E-Commerce AI Assistants", "published": "2026-06-11T07:21:55Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.12940", "title": "Self-Guidance: Enhancing Neural Codecs via Decoder Manifold Alignment", "published": "2026-06-11T06:06:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12896", "title": "PolicyGuard: Towards Test-time and Step-level Adversary (Backdoor) Defense for Reinforcement Learning Agent", "published": "2026-06-11T04:54:28Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12881", "title": "Direct Preference Optimization for Chatbot Fine-Tuning: An Empirical Study", "published": "2026-06-11T04:15:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12876", "title": "Multi-Bitwidth Quantization for LLMs Using Additive Codebooks", "published": "2026-06-11T04:06:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12871", "title": "DailyReport: An Open-ended Benchmark for Evaluating Search Agents on Daily Search Tasks", "published": "2026-06-11T03:59:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12708", "title": "AfriSUD: A Dependency Treebank Collection for Evaluating Models on African Languages", "published": "2026-06-10T21:55:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12578", "title": "MARD: Mirror-Augmented Reasoning Distillation for Mechanism-Level Drug-Drug Interaction Prediction", "published": "2026-06-10T18:26:11Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12360", "title": "Anatomy of Post-Training: Using Interpretability to Characterize Data and Shape the Learning Signal", "published": "2026-06-10T17:31:16Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12299", "title": "Learning What to Say to Your VLA: Mostly Harmless Vision Language Action Model Steering", "published": "2026-06-10T16:34:49Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12281", "title": "CCKS: Consensus-based Communication and Knowledge Sharing", "published": "2026-06-10T16:20:18Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.12243", "title": "VIA-SD: Verification via Intra-Model Routing for Speculative Decoding", "published": "2026-06-10T15:45:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12203", "title": "Adaptive Multi-Resolution Procedural Knowledge Compression for Large Language Models", "published": "2026-06-10T15:21:18Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12198", "title": "LLM-Based User Personas for Recommendations at Scale", "published": "2026-06-10T15:18:32Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "llm-recommendation", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.11953", "title": "Decoding Multimodal Cues: Unveiling the Implicit Meaning Behind Hateful Videos", "published": "2026-06-10T11:28:23Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11925", "title": "Corpus Augmentation for Sign Language Translation via LLM-Guided Video Stitching", "published": "2026-06-10T10:56:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.12487", "title": "DynamicPTQ: Mitigating Activation Quantization Collapse via Residual-Stream Dynamics", "published": "2026-06-10T09:25:45Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11709", "title": "RLCSD: Reinforcement Learning with Contrastive On-Policy Self-Distillation", "published": "2026-06-10T06:31:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11680", "title": "Organize then Retrieve: Hierarchical Memory Navigation for Efficient Agents", "published": "2026-06-10T05:49:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11652", "title": "IAPO: Input Attribution-Aware Policy Optimization for Tool Use in Small Multimodal Agents", "published": "2026-06-10T04:30:37Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11625", "title": "TimeRouter: Efficient and Adaptive Routing of Time-Series Foundation Models", "published": "2026-06-10T03:39:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11616", "title": "DeMix: Debugging Training Data with Mixed Data Error Types by Investigating Influence Vectors", "published": "2026-06-10T03:28:17Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11560", "title": "LLMs+Graphs: Toward Graph-Native, Synergistic AI Systems", "published": "2026-06-10T01:39:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11552", "title": "Teaching Diffusion to Speculate Left-to-Right", "published": "2026-06-10T01:21:56Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11542", "title": "Pretrained self-supervised speech models can recognize unseen consonants", "published": "2026-06-10T01:07:32Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11512", "title": "SAGE: Answer-Conditioned Uncertainty Targets for Verbal Uncertainty Alignment", "published": "2026-06-09T23:17:16Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11499", "title": "Hubs or Fringes: Pretraining Data Selection via Web Graph Centrality", "published": "2026-06-09T22:44:47Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11437", "title": "The Power of Test-Time Training for Approximate Sampling", "published": "2026-06-09T20:48:48Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11382", "title": "GLACIER: A Multimodal Student-Teacher Foundation Model for Molecular Property Prediction", "published": "2026-06-09T19:05:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11350", "title": "When More Documents Hurt RAG: Mitigating Vector Search Dilution with Domain-Scoped, Model-Agnostic Retrieval", "published": "2026-06-09T18:26:24Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11337", "title": "Can AI Agents Synthesize Scientific Conclusions?", "published": "2026-06-09T18:16:04Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11087", "title": "Test-Time Gradient Guidance of Flow Policies in Reinforcement Learning", "published": "2026-06-09T16:45:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11078", "title": "A History-Aware Visually Grounded Critic for Computer Use Agents", "published": "2026-06-09T16:39:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11033", "title": "AuRA: Internalizing Audio Understanding into LLMs as LoRA", "published": "2026-06-09T16:05:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11023", "title": "Generative Archetype-Grounded Item Representations for Sequential Recommendation", "published": "2026-06-09T15:59:14Z", "tracks": [ "foundation-model", "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10918", "title": "Task Robustness via Re-Labelling Vision-Action Robot Data", "published": "2026-06-09T14:28:22Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10907", "title": "From Prompt to Purchase: How AI Brand Recommendations Move Consumers on the Open Web", "published": "2026-06-09T14:16:48Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google", "priority-org-netflix" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.10860", "title": "Training LLMs to Enforce Multi-Level Instruction Hierarchies via Gravity-Weighted Direct Preference Optimization", "published": "2026-06-09T13:39:17Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10771", "title": "On-sky demonstration of reinforcement learning for adaptive optics control", "published": "2026-06-09T12:26:06Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10768", "title": "N-GRPO: Embedding-Level Neighbor Mixing for Enhanced Policy Optimization", "published": "2026-06-09T12:21:27Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10716", "title": "Attention Expansion: Enhancing Keyphrase Extraction from Long Documents with Attention-Augmented Contextualized Embeddings", "published": "2026-06-09T11:24:07Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10706", "title": "Unifying Data, Memory, and Compute Efficiency in LLM training: A Survey", "published": "2026-06-09T11:09:58Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10581", "title": "ParaBridge: Bridging Paralinguistic Perception and Dialogue Behavior in Speech Language Models", "published": "2026-06-09T08:45:52Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10487", "title": "Stop Early, Spend Less: Hidden-State Probes as a Practical Recipe for Streaming Moderation of LLM Outputs", "published": "2026-06-09T07:01:43Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10481", "title": "Advancing the State-of-the-Art in Empirical Privacy Auditing", "published": "2026-06-09T06:50:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10466", "title": "UPLOTS: A Unified Pretrained Language Model for Constrained Time-series Generation", "published": "2026-06-09T06:36:06Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10423", "title": "WebChallenger: A Reliable and Efficient Generalist Web Agent", "published": "2026-06-09T04:53:19Z", "tracks": [ "agent", "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.24768", "title": "PATHFinder Agent for Tailored Prenatal Care", "published": "2026-06-09T03:40:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10369", "title": "PADD: Path-Aligned Decompression Distillation for Non-Router Teacher to Guide MoE Student Learning", "published": "2026-06-09T03:28:17Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10357", "title": "Atomic Intent Reasoning: Bringing LLM Semantics to Industrial Cross-Domain Recommendations", "published": "2026-06-09T03:13:52Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "llm-recommendation", "priority-org-kuaishou", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.10321", "title": "Baseline-Free Policy Optimization for Neural Combinatorial Optimization", "published": "2026-06-09T02:18:30Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11257", "title": "Energy-Efficient On-Device RAG on a Mobile NPU: System Design and Benchmark on Snapdragon X Elite", "published": "2026-06-09T01:09:00Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10267", "title": "What Matters in Orchestrating Robot Policies: A Systematic Study of Hierarchical VLA Agents", "published": "2026-06-09T00:24:00Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10243", "title": "DUET -- Dual User Embedding Transformers for Offsite Conversion Prediction", "published": "2026-06-08T23:13:58Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "recommendation-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.10217", "title": "Alignment Defends LLMs from Property Inference Attacks", "published": "2026-06-08T22:15:20Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10216", "title": "A Source Domain is All You Need: Source-Only Cross-OS Transfer Learning for APT Anomaly Detection via Semantic Alignment and Optimal Transport", "published": "2026-06-08T22:13:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10209", "title": "Less Context, Better Agents: Efficient Context Engineering for Long-Horizon Tool-Using LLM Agents", "published": "2026-06-08T22:01:28Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10156", "title": "$τ$-Rec: A Verifiable Benchmark for Agentic Recommender Systems", "published": "2026-06-08T20:35:45Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10147", "title": "From Senses to Decisions: The Information Flow of Auditory and Visual Perception in Multimodal LLMs", "published": "2026-06-08T20:26:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10126", "title": "Pareto-Guided Teacher Alignment for Fair Personalized Text Generation", "published": "2026-06-08T19:57:13Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10120", "title": "MetaPlate: Counterfactual-Guided RAG-LLM Tool for Personalized Food Recommendation and Hyperglycemia Prevention", "published": "2026-06-08T19:52:08Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.10078", "title": "Mult-DPO: Multinomial Direct Preference Optimization for Recommender Systems", "published": "2026-06-08T18:53:45Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09821", "title": "Rethinking the Divergence Regularization in LLM RL", "published": "2026-06-08T17:58:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09770", "title": "Discovering Functionally Selective Brain Regions with a Deep Topographic Multimodal Model", "published": "2026-06-08T17:31:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09595", "title": "Popcorn: A Configurable Benchmark for Visual Evidence in Multimodal Movie Recommendation", "published": "2026-06-08T15:06:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09590", "title": "Clinically Grounded Privacy Evaluation of Medical LMs", "published": "2026-06-08T15:02:19Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09570", "title": "UXBench: Benchmarking User Experience in AI Assistants", "published": "2026-06-08T14:44:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09525", "title": "Emergence of Context Characteristics Sensitivity in Large Language Models", "published": "2026-06-08T14:11:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09514", "title": "BUDDY: BUdget-Driven DYnamic Depth Routing for Adaptive Large Language Model Inference", "published": "2026-06-08T14:06:35Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09471", "title": "Escaping the KL Agreement Trap in On-Policy Distillation", "published": "2026-06-08T13:28:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09470", "title": "A Finetuned SpeechLLM for Joint Multi-Granular L2 Assessment and Natural-Language Rationales", "published": "2026-06-08T13:27:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09466", "title": "DECSELFMASK: Leveraging Unlabeled Text via Self-Relevance-Guided Masking for Decoder-Only Classification", "published": "2026-06-08T13:21:42Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09456", "title": "Breaking the Tokenizer Barrier: On-Policy Distillation across Model Families", "published": "2026-06-08T13:12:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09401", "title": "Benchmarking Empirical Privacy Protection for Adaptations of Large Language Models", "published": "2026-06-08T12:21:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference", "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09399", "title": "RunAgent SuperBrowser: A Theory of Autonomous Web Navigation Grounded in Human Browsing Behaviour", "published": "2026-06-08T12:18:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09388", "title": "Distilling Safe LLM Systems via Soft Prompts for On Device Settings", "published": "2026-06-08T12:03:51Z", "tracks": [ "foundation-model", "post-training" ], "review_buckets": [ "p2-deferred-review", "query-collision" ], "matched_queries": [ "efficient-inference", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09348", "title": "PBSD: Privileged Bayesian Self-Distillation for Long-Horizon Credit Assignment", "published": "2026-06-08T11:20:58Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09956", "title": "Multi-task LLMs for Bug Classification: Efficient Inference with Auxiliary Decoding Heads", "published": "2026-06-08T11:15:49Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "efficient-inference" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09338", "title": "Multi-Hop Knowledge Composition is Bound by Pretraining Exposure", "published": "2026-06-08T11:05:17Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09204", "title": "The Injection Paradox: Brand-Level Suppression in Safety-Trained LLM Recommendations via RAG Context Injection", "published": "2026-06-08T08:38:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09059", "title": "Stage-1 Controls the Entropy Regime, Not the Outcome", "published": "2026-06-08T05:49:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09052", "title": "INFUSER: Influence-Guided Self-Evolution Improves Reasoning", "published": "2026-06-08T05:40:36Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09043", "title": "DynaCF: Mitigating Shortcut Learning in Reward Models via Dynamic Counterfactual Sensitivity", "published": "2026-06-08T05:24:15Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09033", "title": "CRANE: Knowledge Editing for Reasoning MLLMs", "published": "2026-06-08T05:01:11Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09032", "title": "Bridging the Agent-World Gap: Text World Models for LLM-based Agents", "published": "2026-06-08T04:58:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.19202", "title": "Active Inference as Context Acquisition for AI Agents", "published": "2026-06-08T02:47:35Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08843", "title": "From A to B to A: Palindromic Zero-Shot Voice Conversion with Non-Parallel Data", "published": "2026-06-07T21:25:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.18094", "title": "NE-BERT: A Multilingual Language Model for Nine Northeast Indian Languages", "published": "2026-06-07T17:05:23Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08728", "title": "Artificial Intelligence for Mathematical Reasoning: An Integrated Survey of Language Models, Neuro-symbolic Systems, and Verified Discovery", "published": "2026-06-07T16:50:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08679", "title": "Rank Intervals for Leaderboards: A Hierarchical Framework for Model Evaluation", "published": "2026-06-07T15:31:29Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08656", "title": "From Player to Master: Enhancing Test-Time Learning of LLM Agents via Reinforcement Learning over Memory", "published": "2026-06-07T14:53:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08625", "title": "From Holistic Evaluation to Structured Criteria: Rubrics Across the Evolving LLM Landscape", "published": "2026-06-07T13:34:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08604", "title": "Gryphon: A Unified Architecture for Semantic-ID Generation and Item-Level Scoring in Industrial Recommendations", "published": "2026-06-07T12:31:26Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.08602", "title": "Reinforcement Learning for Flow-Matching Policies with Density Transport", "published": "2026-06-07T12:28:51Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08501", "title": "Back on Track: Aligning Rewards and States for Reasoning in Diffusion Large Language Models", "published": "2026-06-07T07:59:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08496", "title": "SAEExplainer: Interpreting SAE Features with Activation-Guided Preference Optimization", "published": "2026-06-07T07:54:25Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08480", "title": "Adaptive Loss Balancing for Noise-Robust GRPO in Generative Recommendation", "published": "2026-06-07T06:51:18Z", "tracks": [ "post-training", "recommendation" ], "review_buckets": [ "industrial-fulltext-review", "p2-deferred-review" ], "matched_queries": [ "recommendation-ranking", "test-time-rl" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.08466", "title": "ToolRec: Calibrated Preference Alignment for Query Recommendation in On-Device Assistants", "published": "2026-06-07T06:06:38Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "llm-recommendation", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.08446", "title": "Sparrow: Sparse Rollout for Stable and Efficient Long-context RL of Large Language Models", "published": "2026-06-07T04:24:45Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08346", "title": "CATPO: Critique-Augmented Tree Policy Optimization", "published": "2026-06-06T21:29:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08221", "title": "De novo molecular generation with optical property preconditioning at the token level", "published": "2026-06-06T15:16:40Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08169", "title": "CLASP: Language-Driven Robot Skill Selection and Composition using Task-Parameterized Learning", "published": "2026-06-06T13:33:39Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08157", "title": "Cross Paraphrastic Invariance Learning for Hallucination Detection", "published": "2026-06-06T13:13:12Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08147", "title": "Biological Reasoning-Informed Regression for Interpretable Regulatory DNA Activity Prediction", "published": "2026-06-06T12:56:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.08051", "title": "How Small Can You Go? LoRA Fine-Tuning 270M-8B Models for Merchant Information Extraction in Financial Transactions", "published": "2026-06-06T08:32:38Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07996", "title": "MC-PDD: Masked Corpus-Level Pretraining Data Detection for Black-Box Large Language Models", "published": "2026-06-06T06:27:54Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07972", "title": "OneFeed: A Unified Generative Framework for Feed Content Enhancement and Query Generation", "published": "2026-06-06T04:17:09Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07936", "title": "Illusions of the Gold Standard: A Large-scale Analysis of Human Evaluation Protocols for Long-form Text Generation", "published": "2026-06-06T01:55:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07867", "title": "The Cold-Start Safety Gap in LLM Agents", "published": "2026-06-05T21:58:19Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07853", "title": "Beyond English benchmarks: clinical llm evaluation in Brazilian Portuguese", "published": "2026-06-05T21:29:39Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07845", "title": "GRPO Does Not Close the Multi-Agent Coordination Gap", "published": "2026-06-05T21:13:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07778", "title": "Unlocking Latent Value: Taxonomy-Guided Recovery of High-Performing Data from Low-Tier Web Corpora", "published": "2026-06-05T18:43:14Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07367", "title": "Self-evolving LLM agents with in-distribution Optimization", "published": "2026-06-05T15:09:52Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07317", "title": "Gated Bidirectional Linear Attention for Generative Retrieval", "published": "2026-06-05T14:37:02Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07237", "title": "When Large Language Models Fail in Healthcare: Evaluating Sensitivity to Prompt Variations", "published": "2026-06-05T13:07:11Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07190", "title": "From Correctness to Utility: Gain-Based Prefix Evaluation for LLM Reasoning", "published": "2026-06-05T11:56:50Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07082", "title": "On the Geometry of On-Policy Distillation", "published": "2026-06-05T09:20:15Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07068", "title": "Bias in Filter Feature Selection Evaluation: A Meta-Analysis of Datasets, Baselines, and Experimental Design Choices", "published": "2026-06-05T09:06:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.07017", "title": "The Sim-to-Real Gap of Foundation Model Agents: A Unified MDP Perspective", "published": "2026-06-05T08:00:25Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06970", "title": "SSRLive: Live Streaming Recommendation with Dynamic Semantic ID", "published": "2026-06-05T06:58:07Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.06947", "title": "DREAM: Dynamic Refinement of Early Assignment Mappings", "published": "2026-06-05T06:21:01Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07688", "title": "TRACER: Token ReAssignment for Concept ERasure in Generative Recommendation", "published": "2026-06-05T05:19:31Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06892", "title": "GRASP: Geometry-aware Residual Alignment for Scalable Pretraining Data Attribution", "published": "2026-06-05T04:17:50Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06888", "title": "Data-Constrained Language Model Pretraining: Improved Regularization and Scaling Laws", "published": "2026-06-05T04:10:09Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.22574", "title": "Too much evidence, too little time: From text to actionable recommendations through multi-objective evidence reasoning", "published": "2026-06-05T03:53:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06835", "title": "Translate-R1: Cost-Aware Translation Tool Use via Reinforcement Learning", "published": "2026-06-05T02:21:41Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06828", "title": "AdaGRPO: A Capability-Aware Adaptive Enhancement for Flow-based GRPO", "published": "2026-06-05T02:07:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06820", "title": "SCALE: Scalable Cross-Attention Learning with Extrapolation for Agentic Workflow Scheduling", "published": "2026-06-05T01:45:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06779", "title": "Mind the Gap: Bridging Behavioral Silos with LLMs in Multi-Vertical Recommendations", "published": "2026-06-04T23:39:52Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06741", "title": "OpenSkill: Open-World Self-Evolution for LLM Agents", "published": "2026-06-04T21:55:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06708", "title": "Signal-Driven Observation for Long-Horizon Web Agents", "published": "2026-06-04T20:48:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07678", "title": "DOG-DPO:Dynamic Optimization in Geometry for Safety Alignment", "published": "2026-06-04T20:23:23Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06586", "title": "PolyFact: Comparing Consistency-Driven Post-training Methods for Cross-Lingual Factual Recall", "published": "2026-06-04T18:00:02Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06447", "title": "Latent Reasoning with Normalizing Flows", "published": "2026-06-04T17:44:17Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06428", "title": "Reinforcement Learning Elicits Contextual Learning of Unseen Language Translation", "published": "2026-06-04T17:32:06Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06345", "title": "Boosting Brain-to-Image Decoding with TRIBE v2 Data Augmentation", "published": "2026-06-04T16:18:08Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06333", "title": "Subspace-Aware Sparse Autoencoders for Effective Mechanistic Interpretability", "published": "2026-06-04T16:08:25Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06260", "title": "OneReason Technical Report", "published": "2026-06-04T15:04:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06178", "title": "Learning to Route LLMs from Implicit Cost-Performance Preferences via Meta-Learning", "published": "2026-06-04T13:53:03Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06133", "title": "TLA-Prover: Verifiable TLA+ Specification Synthesis via Preference-Optimized Low-Rank Adaptation", "published": "2026-06-04T13:17:06Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.05920", "title": "Asuka-Bench: Benchmarking Code Agents on Underspecified User Intent and Multi-Round Refinement", "published": "2026-06-04T09:24:30Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.06546", "title": "Elmes*: Automated Construction of Fine-Grained Evaluation Rubrics for Large Language Models in Long-Tail Educational Scenarios", "published": "2026-06-04T07:40:12Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.05734", "title": "When AI Says It Feels", "published": "2026-06-04T05:49:34Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.05621", "title": "ANCHOR: Agentic Noise Creation Framework for Human Simulation and Denoising Recommendation", "published": "2026-06-04T02:46:31Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.05597", "title": "AsyncWebRL: Efficient Asynchronous Reinforcement Learning for Multi-Step Visual Web Agents", "published": "2026-06-04T02:18:44Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.05436", "title": "Ten Headache Specialists versus Artificial Intelligence for Clinical Literature Summarization: A Critical Evaluation and Comparison", "published": "2026-06-03T20:58:43Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.05336", "title": "Self-supervised User Profile Generation for Personalization", "published": "2026-06-03T18:25:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.05296", "title": "Agentic Monte Carlo: Simulating Reinforcement Learning for Black-Box Agents", "published": "2026-06-03T18:00:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.05152", "title": "Reinforcement Learning from Rich Feedback with Distributional DAgger", "published": "2026-06-03T17:54:04Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.05257", "title": "Scaling Laws for Behavioral Foundation Models over User Event Sequences", "published": "2026-06-03T15:59:25Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.05253", "title": "Alpha-RTL: Test-Time Training for RTL Hardware Optimization", "published": "2026-06-03T14:51:33Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04928", "title": "Data Attribution in Large Language Models via Bidirectional Gradient Optimization", "published": "2026-06-03T14:21:53Z", "tracks": [ "foundation-model" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "pretraining-data" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04889", "title": "GRAIL: Gradient-Reweighted Advantages for Reinforcement Learning with Verifiable Rewards", "published": "2026-06-03T13:51:27Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04815", "title": "Learning While Acting: A Skill-Enhanced Test-Time Co-Evolution Framework for Online Lifelong Learning Agents", "published": "2026-06-03T12:38:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04807", "title": "BiasGRPO: Stabilizing Bias Mitigation in High-Variance Reward Landscapes via Group-Relative Policy Optimization", "published": "2026-06-03T12:31:42Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04781", "title": "AIP: A Graph Representation for Learning and Governing Agent Skills", "published": "2026-06-03T12:02:49Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04727", "title": "EviRank: Evidence-Based Confidence Estimation for LLM-Based Ranking", "published": "2026-06-03T11:11:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04703", "title": "Rethinking Continual Experience Internalization for Self-Evolving LLM Agents", "published": "2026-06-03T10:30:09Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04694", "title": "DuDi: Dual-Signal Distillation with Cross-Lingual Verbalizer", "published": "2026-06-03T10:23:05Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09887", "title": "SocraticPO: Policy Optimization via Interactive Guidance", "published": "2026-06-03T09:08:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04550", "title": "Trading Engagement for Sustainability: Carbon-Aware Re-ranking for E-commerce Recommendations", "published": "2026-06-03T07:34:46Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04547", "title": "Beyond Retrieval: Learning Compact User Representations for Scalable LLM Personalization", "published": "2026-06-03T07:32:18Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04514", "title": "SAILRec: Steering LLM Attention to Dual-Side Semantically Aligned Collaborative Embeddings for Recommendation", "published": "2026-06-03T06:46:32Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.09883", "title": "TD-Grokking: Learning from Zero-Reward Problems by Training-Time Decomposition", "published": "2026-06-03T06:40:04Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04507", "title": "Self-Evolving Deep Research via Joint Generation and Evaluation", "published": "2026-06-03T06:38:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04455", "title": "The Meta-Agent Challenge: Are Current Agents Capable of Autonomous Agent Development?", "published": "2026-06-03T04:58:17Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04448", "title": "Bridging Short Videos and Live Streams: Reasoning-Guided Multimodal LLMs for Cross-Domain Representation Learning", "published": "2026-06-03T04:49:01Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "llm-recommendation", "priority-org-kuaishou" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.04396", "title": "Read the Trace, Steer the Path: Trajectory-Aware Reinforcement Learning for Diffusion Language Models", "published": "2026-06-03T03:22:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04387", "title": "Rethinking Sales Lead Scoring with LLM-based Hierarchical Preference Ranking", "published": "2026-06-03T03:05:57Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "llm-recommendation", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.05233", "title": "Domain-Conditioned Safety in Frontier Computer-Using Agents: A 793-Episode Browser Benchmark, a Coding-Domain Cross-Reference, and a Reproducibility Audit of Recent Red-Teaming", "published": "2026-06-03T01:21:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04325", "title": "Parameter-Efficient Fine-Tuning with Learnable Rank", "published": "2026-06-03T00:57:36Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04284", "title": "Sparse Mixture-of-Experts Reward Models Learn Interpretable and Specialized Experts for Personalized Preference Modeling", "published": "2026-06-02T23:19:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04246", "title": "StepPRM-RTL: Stepwise Process-Reward Guided LLM Fine-Tuning for Enhanced RTL Synthesis", "published": "2026-06-02T21:52:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04120", "title": "SaliMory: Orchestrating Cognitive Memory for Conversational Agents", "published": "2026-06-02T18:31:50Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04110", "title": "Variance Reduction for Heavy-Tailed Monetization Metrics in Ranking Experiments via Post-Stratification", "published": "2026-06-02T18:14:14Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.04095", "title": "POLARIS: Guiding Small Models to Write Long Stories", "published": "2026-06-02T18:00:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03979", "title": "Language Models Need Sleep: Learning to Self-Modify and Consolidate Memories", "published": "2026-06-02T17:56:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03965", "title": "Agentic Chain-of-Thought Steering for Efficient and Controllable LLM Reasoning", "published": "2026-06-02T17:51:30Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03962", "title": "Using Reward Uncertainty to Induce Diverse Behaviour in Reinforcement Learning", "published": "2026-06-02T17:50:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03866", "title": "Taiji: Pareto Optimal Policy Optimization with Semantics-IDs Trade-off for Industrial LLM-Enhanced Recommendation", "published": "2026-06-02T16:39:06Z", "tracks": [ "post-training", "recommendation" ], "review_buckets": [ "industrial-fulltext-review", "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "preference-optimization", "priority-org-kuaishou", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.04075", "title": "Large Language Models Hack Rewards, and Society", "published": "2026-06-02T16:29:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03810", "title": "Consistency Training Can Entrench Misalignment", "published": "2026-06-02T15:54:24Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03782", "title": "Reasoning over Grammar: Can Synthetic Linguistic Reasoning Traces Enhance Low-Resource Machine Translation?", "published": "2026-06-02T15:36:12Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04071", "title": "Covert Influence Between Language Models", "published": "2026-06-02T15:10:56Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03647", "title": "Black-box, Adaptive, Efficient, Transferable, Harmful, Applicable... Attacks Are All You Need to Break LLMs", "published": "2026-06-02T13:39:15Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03608", "title": "Exploiting Verification-Generation Gap: Test-Time Reinforcement Learning with Confidence-Conditioned Verification", "published": "2026-06-02T13:11:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03603", "title": "World Models Meet Language Models: On the Complementarity of Concrete and Abstract Reasoning", "published": "2026-06-02T13:07:49Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03376", "title": "P$^2$-DPO: Grounding Hallucination in Perceptual Processing via Calibration Direct Preference Optimization", "published": "2026-06-02T09:22:53Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03327", "title": "CAPER: Clause-Aligned Process Supervision for Text-to-SQL", "published": "2026-06-02T08:35:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03221", "title": "VirtualMLE: A Virtual ML Engineer that Optimizes Sequential Recommenders", "published": "2026-06-02T06:31:15Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation", "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03128", "title": "Decoupled Smart Contract Audits: Lightweight LLM Framework via Distillation and Aggregation", "published": "2026-06-02T04:13:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03102", "title": "Small RL Controller, Large Language Model: RL-Guided Adaptive Sampling for Test-Time Scaling", "published": "2026-06-02T03:42:04Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03089", "title": "Constitutional On-Policy Safe Distillation", "published": "2026-06-02T03:17:56Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.04036", "title": "Self-Distilled Policy Gradient", "published": "2026-06-02T02:31:13Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.03021", "title": "Hint-Guided Diversified Policy Optimization for LLM Reasoning", "published": "2026-06-02T01:55:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.28344", "title": "PIXELRAG: Web Screenshots Beat Text for Retrieval-Augmented Generation", "published": "2026-06-01T23:20:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.02883", "title": "LLM-Assisted Reranking to Operationalize Nuanced Objectives in Recommender Systems", "published": "2026-06-01T20:54:14Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.02812", "title": "Traj-Evolve: A Self-Evolving Multi-Agent System for Patient Trajectory Modeling in Lung Cancer Early Detection", "published": "2026-06-01T19:30:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.02741", "title": "Greener Than Humans? Environmental Attitudes in Large Language Models", "published": "2026-06-01T18:05:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.02684", "title": "Filter, Then Reweight: Rethinking Optimization Granularity in On-Policy Distillation", "published": "2026-06-01T17:58:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.02530", "title": "SafeSteer: Localized On-Policy Distillation for Efficient Safety Alignment", "published": "2026-06-01T17:38:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.02404", "title": "K-BrowseComp: A Web Browsing Agent Benchmark Grounded in Korean Contexts", "published": "2026-06-01T15:50:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.02373", "title": "Harness-1: Reinforcement Learning for Search Agents with State-Externalizing Harnesses", "published": "2026-06-01T15:21:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.02372", "title": "COMAP: Co-Evolving World Models and Agent Policies for LLM Agents", "published": "2026-06-01T15:21:17Z", "tracks": [ "agent", "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.02339", "title": "Entropy Minimization without Model Collapse: Mitigating Prediction Bias in Medical Imaging", "published": "2026-06-01T14:48:16Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.02255", "title": "Who Annotates in NLP? A Large-scale Assessment of Human Annotation Reporting between 2018 and 2025", "published": "2026-06-01T13:43:10Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "llm-recommendation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12330", "title": "Reliability-Aware Sexism Detection: Combining DPO with Annotator Agreement and Token-Level Confidence Scoring", "published": "2026-06-01T12:35:41Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.02031", "title": "OpenWebRL: Demystifying Online Multi-turn Reinforcement Learning for Visual Web Agents", "published": "2026-06-01T10:20:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01993", "title": "MMG2Skill: Can Agents Distill In-the-Wild Guides into Self-Evolving Skills?", "published": "2026-06-01T09:50:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01934", "title": "HMPO: Hybrid Median-length Policy Optimization for Chain-of-Thought Compression", "published": "2026-06-01T09:01:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01783", "title": "Breaking the Information Silo: Semantic Personas for Cross-Domain Recommendation", "published": "2026-06-01T07:04:10Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01755", "title": "TriAlign: Towards Universal Truth Consistency in Personalized LLM Alignment", "published": "2026-06-01T06:19:52Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01682", "title": "Off-the-Shelf LLMs as Process Scorers: Training-Free Alternative to PRMs for Mathematical Reasoning", "published": "2026-06-01T04:43:36Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01635", "title": "AlphaToken: Decoupling Adaptation and Stability for Path-Aware Response Token Valuation in LLM Post-Training", "published": "2026-06-01T03:40:35Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01561", "title": "S-SPPO: Semantic-Calibrated Self-Play Preference Optimization", "published": "2026-06-01T02:06:58Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01533", "title": "Multi-Agent Computer Use", "published": "2026-06-01T01:29:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01476", "title": "OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification", "published": "2026-05-31T22:31:15Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01456", "title": "Truthful AI Advisors: A Pre-Specified Benchmark for Large Language Model Honesty Under Preference Misalignment", "published": "2026-05-31T21:30:43Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01382", "title": "Efficient Exploration for Iterative Nash Preference Optimization", "published": "2026-05-31T18:11:26Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.02643", "title": "Inference Cost Attacks for Retrieval-Augmented Large Language Models", "published": "2026-05-31T15:11:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01249", "title": "Trust Region On-Policy Distillation", "published": "2026-05-31T14:04:51Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2608.12327", "title": "Comparative Analysis of Multilingual Pre-trained Models for Nepali Automatic Speech Recognition", "published": "2026-05-31T12:05:48Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01111", "title": "LeAP: Learnable Adaptive Permutation for Feature Selection in Heterogeneous and Sparse Recommender Systems", "published": "2026-05-31T09:12:55Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01091", "title": "Deep Research as Rubric for Reinforcement Learning", "published": "2026-05-31T08:25:04Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.01039", "title": "OPD+: Rethinking the Advantage Design for On-Policy Distillation", "published": "2026-05-31T06:10:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00931", "title": "CV-Arena: An Open Benchmark for Instructional Computer Vision Problem Solving with Human-AI Collaborative Preferences", "published": "2026-05-30T23:37:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07629", "title": "Large Language Models Should Learn Personalized Rather Than Aggregated Human Preferences", "published": "2026-05-30T18:47:52Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00755", "title": "Internalize the Temperature: On-Policy Self-Distillation as Policy Reheater for Reinforcement Learning", "published": "2026-05-30T14:44:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00750", "title": "I-WebGenBench : Evaluating Interactivity in LLM-Generated Scientific Web Applications", "published": "2026-05-30T14:34:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.20482", "title": "PersonaTrail: Benchmarking Personalized Web Agents through Browsing Trails", "published": "2026-05-30T13:27:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00674", "title": "The Paradox of Outcome Optimization: A Causal Information-Theoretic Bound on Reasoning Shortcuts in LLMs", "published": "2026-05-30T11:06:13Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00579", "title": "Sandboxed Coding Agents are Competitive Omni-modal Task Solvers", "published": "2026-05-30T07:04:27Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00564", "title": "Decomposed On-Policy Distillation for Vision-Language Reasoning: Steering Gradients for Visual Grounding", "published": "2026-05-30T06:34:37Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00547", "title": "Learning to Retrieve: Dual-Level Long-Term Memory for Text-to-SQL Agents", "published": "2026-05-30T05:44:34Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00497", "title": "\"I Strongly Suspect This Website Is a Scam\": Benchmarking PII Leakage and Detection without Defense in Autonomous Web Agents", "published": "2026-05-30T03:00:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00437", "title": "EST-PRM: Stress-Testing Process Reward Models Before They Become Load-Bearing", "published": "2026-05-30T00:05:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00422", "title": "UniPinRec: Unifying Generative Retrieval and Ranking at Pinterest Scale", "published": "2026-05-29T23:17:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence", "priority-org-pinterest" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00408", "title": "Masking Stale Observations Helps Search Agents -- Until It Doesn't: A Regime Map and Its Mechanism", "published": "2026-05-29T22:51:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00376", "title": "The Deterministic Horizon: When Extended Reasoning Fails and Tool Delegation Becomes Necessary", "published": "2026-05-29T21:35:23Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00324", "title": "LLMs Need Encoders for Semantic IDs Too", "published": "2026-05-29T20:01:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-pinterest" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00305", "title": "Bridging Reasoning Trajectories in On-Policy Distillation via Near-Future Guidance", "published": "2026-05-29T19:32:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00282", "title": "Synthetic Data from Cross-Domain Events for Large-Scale Recommendation Systems", "published": "2026-05-29T19:17:50Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.31584", "title": "LongTraceRL: Learning Long-Context Reasoning from Search Agent Trajectories with Rubric Rewards", "published": "2026-05-29T17:51:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.31365", "title": "Learning to Adapt: Self-Improving Web Agent via Cognitive-Aware Exploration", "published": "2026-05-29T14:37:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.31312", "title": "Learning from Fine-Grained Visual Discrepancies: Mitigating Multimodal Hallucinations via In-Context Visual Contrastive Optimization", "published": "2026-05-29T13:44:56Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.31268", "title": "Mellum2 Technical Report", "published": "2026-05-29T13:01:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.31073", "title": "ConsisGuard: Aligning Safety Deliberation with Policy Enforcement in LLM Guardrails", "published": "2026-05-29T09:42:08Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07603", "title": "MetaEvo: A Meta-Optimization Framework for Experience-Driven Agent Evolution", "published": "2026-05-29T09:31:39Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07602", "title": "Sample-Efficient Post-Training for LEGO Spatial-Physics Reasoning", "published": "2026-05-29T09:31:25Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.30912", "title": "Attend to Evidence: Evidence-Anchored Spatial Attention Supervision for Multimodal RLVR", "published": "2026-05-29T06:50:46Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.30848", "title": "LLM Anonymization Against Agentic Re-Identification", "published": "2026-05-29T05:12:39Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.30833", "title": "Your Teacher Can't Help You Here: Combating Supervision Fidelity Decay in On-Policy Distillation", "published": "2026-05-29T04:39:20Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.30776", "title": "Efficient and Uncertainty-Aware Diffusion Framework for Offline-to-Online Reinforcement Learning", "published": "2026-05-29T03:08:42Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.30758", "title": "Pairwise Reference Alignment as a Model-Level Ordinal Observable", "published": "2026-05-29T02:41:18Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.30727", "title": "MosaicLeaks:Privacy Risks in Querying-in-the-Open for Deep Research Agents", "published": "2026-05-29T01:44:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.30524", "title": "Representation Collapse in Sequential Post-Training of Large Language Models", "published": "2026-05-28T19:59:17Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.30504", "title": "Auditing LLM Benchmarks with Item Response Theory", "published": "2026-05-28T19:38:44Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.30274", "title": "Loong: A Human-Like Long Document Translation Agent with Observe-and-Act Adaptive Context Selection", "published": "2026-05-28T17:32:25Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.30273", "title": "LLUMI: Improving LLM Writing Assistance for Mental Health Support with Online Community Feedback", "published": "2026-05-28T17:30:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.30251", "title": "Same Evidence, Different Answers: Canonical-Context On-Policy Distillation for Multi-Turn Language Models", "published": "2026-05-28T17:14:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.30169", "title": "Dissociative Identity: Language Model Agents Lack Grounding for Reputation Mechanisms", "published": "2026-05-28T16:20:19Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29888", "title": "LaRA: Layer-wise Representation Analysis for Detecting Data Contamination in RL Post-Training", "published": "2026-05-28T13:13:49Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29886", "title": "CRITIC-R1: Learning Structured Critics for Retrieval-Augmented Generation", "published": "2026-05-28T13:11:36Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29861", "title": "Towards Verifiable Multimodal Deep Research: A Multi-Agent Harness for Interleaved Report Generation", "published": "2026-05-28T12:40:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29860", "title": "ESPO: Early-Stopping Proximal Policy Optimization", "published": "2026-05-28T12:40:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29648", "title": "Verifiable Rewards Beyond Math and Code: Lightweight Corpus-Grounded Process Supervision for Factual Question Answering", "published": "2026-05-28T09:14:37Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29615", "title": "DiffSpot: Can VLMs Spot Fine-Grained Visual Differences in Web Interfaces?", "published": "2026-05-28T08:50:34Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29584", "title": "GAPD: Gold-Action Policy Distillation for Agentic Reinforcement Learning in Knowledge Base Question Answering", "published": "2026-05-28T08:28:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29559", "title": "LiteCoder-Terminal: Scaling Long-Horizon Terminal Environments for Learning Language Agents", "published": "2026-05-28T08:11:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29495", "title": "On-Policy Replay for Continual Supervised Fine-Tuning", "published": "2026-05-28T07:19:47Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29475", "title": "MOOSE-Copilot: A Web-Based Interactive Assistant for Unified Exploratory and Fine-Grained Scientific Hypothesis Discovery", "published": "2026-05-28T07:06:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.00017", "title": "Learning User-Aware Recall: Personalized Retrieval in Long-Term Conversational Memory", "published": "2026-05-28T06:47:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29398", "title": "GDSD: Reinforcement Learning as Guided Denoiser Self-Distillation for Diffusion Language Models", "published": "2026-05-28T05:47:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29368", "title": "SURGENT: A Surgical Multi-Agent Assistance System Across the Perioperative Workflow", "published": "2026-05-28T05:12:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29343", "title": "Draft-OPD: On-Policy Distillation for Speculative Draft Models", "published": "2026-05-28T04:30:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29336", "title": "Enhancing Factuality through Consensus and Consistency in Summarization Using Minimum Bayes Risk Decoding", "published": "2026-05-28T04:14:43Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29310", "title": "Rubric-Guided Process Reward for Stepwise Model Routing", "published": "2026-05-28T03:42:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29287", "title": "UniNote: A Unified Embedding Model for Multimodal Representation and Ranking", "published": "2026-05-28T03:11:54Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29224", "title": "Relevance as a Vulnerability: How Web Retrieval Degrades Safety Alignment in LLM Agents", "published": "2026-05-28T01:23:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29218", "title": "GTA: Generating Long-Horizon Tasks for Web Agents at Scale", "published": "2026-05-28T01:05:50Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.29141", "title": "Toward User Preference Alignment in LLM Recommendation via Explicit Context Feedback", "published": "2026-05-27T22:10:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.29089", "title": "OISD: On-Policy Internal Self-Distillation of Language Models", "published": "2026-05-27T20:43:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.28918", "title": "When LLM Reward Design Fails: Diagnostic-Driven Refinement for Sparse Structured RL", "published": "2026-05-27T17:57:43Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.28802", "title": "Human Label Variation as Stable Signal: Learning Annotator-Specific Explanation Behavior via Cross-Annotator Preference Optimization", "published": "2026-05-27T17:55:00Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.28787", "title": "Do Data Agents Need Semantic Metadata? A Comparative Study in Agentic Data Retrieval", "published": "2026-05-27T17:46:43Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.28910", "title": "Hallucination Detection-Guided Preference Optimization for Clinical Summarization", "published": "2026-05-27T17:24:26Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.28629", "title": "Mobile-Aptus: Confidence-Driven Proactive and Robust Interaction in MLLM-based Mobile-Using Agents", "published": "2026-05-27T15:37:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.28440", "title": "AdaDPO: Self-Adaptive Direct Preference Optimization with Balanced Gradient Updates", "published": "2026-05-27T13:05:49Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.28396", "title": "ADWIN: Adaptive Windows for Horizon-Aware On-Policy Distillation", "published": "2026-05-27T12:33:44Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.28287", "title": "AtomComposer: Discovering Chemical Space from First Principles with Reinforcement Learning", "published": "2026-05-27T10:35:28Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.28888", "title": "Generative Spatiotemporal Intent Sequence Recommendation via Implicit Reasoning in Amap", "published": "2026-05-27T07:27:32Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-alibaba", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.28023", "title": "VCap: Hypergeometric Rewards for Weak-to-Strong Visual Captioning", "published": "2026-05-27T06:27:04Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.28017", "title": "Can It Reach the Generator? Investigating the Survival of Prompt-Injection Attacks in Realistic RAG Settings", "published": "2026-05-27T06:16:11Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.28014", "title": "ROSD: Reflective On-Policy Self-Distillation for Language Model Reasoning across Domains", "published": "2026-05-27T06:09:29Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27986", "title": "An Evolutionary Approach for Designing Stable and Highly Expressible Low-Immunogenicity Therapeutic mRNA Sequences", "published": "2026-05-27T05:20:17Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27916", "title": "OphIn-500K: Curating Web-Scale Visual Instructions for Scaling Ophthalmic Multimodal Large Language Models", "published": "2026-05-27T03:43:23Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27881", "title": "Retrieval, Reward, and Training Protocols: What Matters in Training Search Agents?", "published": "2026-05-27T03:04:36Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27873", "title": "AIBuildAI-2: A Knowledge-Enhanced Agent for Automatically Building AI Models", "published": "2026-05-27T02:44:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27858", "title": "DecomposeRL: Learning to Ask Useful, Informative, and Diverse Questions for Semi-Supervised, Traceable Claim Verification", "published": "2026-05-27T02:19:52Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27772", "title": "Do Audio LLMs Listen or Read? Analyzing and Mitigating Paralinguistic Failures with VoxParadox", "published": "2026-05-26T23:44:23Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07583", "title": "Outage Detection in Self-Healing Smart Grids Using Reinforcement Learning with Spectral Graph Neural Networks", "published": "2026-05-26T23:38:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27765", "title": "Restoring the Sweet Spot: Pass-Rate Weighted Self-Distillation for LLM Reasoning", "published": "2026-05-26T23:30:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27736", "title": "Explicit Critic Guidance for Aligning Diffusion Models", "published": "2026-05-26T22:20:51Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27704", "title": "Joint Optimization of Relevance and Engagement in Multi-Task Ranking for E-Commerce with Efficient LLM Supervision", "published": "2026-05-26T21:26:17Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "search-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.27656", "title": "Developing an Intelligent Job Recommendation System Using Semantic Retrieval and Explainable AI Techniques", "published": "2026-05-26T20:16:42Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27567", "title": "Why LLMs Fail at Causal Discovery and How Interventional Agents Escape", "published": "2026-05-26T18:37:03Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27355", "title": "Alignment Tampering: How Reinforcement Learning from Human Feedback Is Exploited to Optimize Misaligned Biases", "published": "2026-05-26T17:57:04Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27255", "title": "Pair-In, Pair-Out: Latent Multi-Token Prediction for Efficient LLMs", "published": "2026-05-26T16:31:45Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27186", "title": "MAIGO: Mitigating Lost-in-Conversation with History-Cleaned On-Policy Self-Distillation", "published": "2026-05-26T15:38:46Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.19357", "title": "Stochastic Primal-Dual Decoding for Multiobjective Generative Recommender Systems", "published": "2026-05-26T15:05:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.27103", "title": "MuChator: Enabling Active Music Discovery via Conversational Music LLMs in Douyin Music", "published": "2026-05-26T14:42:52Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-bytedance", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.27095", "title": "Adversarial Dual On-Policy Distillation from Expressive Teacher", "published": "2026-05-26T14:38:03Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27066", "title": "Large Language Model-Powered Query-Driven Event Timeline Summarization in Industrial Search", "published": "2026-05-26T14:16:27Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27043", "title": "Causal Representation Learning for Generalisable Recommendation", "published": "2026-05-26T13:58:36Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "production-evidence", "recommendation-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.26940", "title": "Accountable Human-AI Deliberation with LLMs: Scaling Collective Intelligence through Symbiotic Scaffolding", "published": "2026-05-26T12:31:37Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.26924", "title": "Learning to Adapt SFT Data for Better Reasoning Generalization", "published": "2026-05-26T12:20:53Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.26849", "title": "Uncertainty-Aware Budget Allocation for Adaptive Test-Time Reasoning", "published": "2026-05-26T11:06:58Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.26785", "title": "EmoDistill: Offline Emotion Skill Distillation for Language Model Agents in Adversarial Negotiation", "published": "2026-05-26T09:54:53Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd", "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.26717", "title": "L2Rec: Towards Dual-View Understanding of LLMs for Personalized Recommendation", "published": "2026-05-26T08:57:19Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B03", "plan_reason": "6–5 月序列建模、生成搜索与多场景排序 — completed" }, { "arxiv_id": "2605.26554", "title": "Linear and Neural Dueling Bandits with Delayed Feedback", "published": "2026-05-26T05:07:25Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.26533", "title": "A Hybrid Vision-Language Architecture for Automated Defect Reasoning and Report Generation in Industrial Inspection", "published": "2026-05-26T04:27:38Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28642", "title": "ThinkReset: Learnable Intermediate Interface Construction for Bounded-Context Long-Horizon Reasoning", "published": "2026-05-26T02:25:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.26424", "title": "Uniboost: Global Coordination with Value Alignment for Fair and Efficient Traffic Allocation", "published": "2026-05-26T01:22:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.26396", "title": "Advancing Creative Physical Intelligence in Large Multimodal Models", "published": "2026-05-25T23:59:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.26385", "title": "Credit-assigned Policy Gradient for Early Stage Retrieval in Two-stage Ranking", "published": "2026-05-25T23:17:37Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recommendation-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.26293", "title": "CroCo: Cross-Lingual Contrastive Preference Tuning on Self-Generations", "published": "2026-05-25T19:30:45Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.26246", "title": "The Bridge-Garden Dilemma in LLM Distillation: Why Mixing Hard and Soft Labels Works", "published": "2026-05-25T18:12:54Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.26111", "title": "Squeezing Capacity from Multimodal Large Language Models for Subject-driven Generation", "published": "2026-05-25T17:59:35Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.26037", "title": "Peak-Then-Collapse and the Four Interface Channels of Knowledge-Graph Tool Use", "published": "2026-05-25T17:05:35Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.25988", "title": "What Makes a Medical Checker Trainable? Diagnosing Signal Collapse and Reward Hacking in Checker-Guided RAG for Biomedical QA", "published": "2026-05-25T16:06:06Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.25966", "title": "Mapping the Schedule x Bit-Width Boundary in Sub-100M Quantisation-Aware Training", "published": "2026-05-25T15:42:34Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.25920", "title": "Can LLMs Time Travel? Enhancing Temporal Consistency in Legal Agentic Search through Reinforcement Learning", "published": "2026-05-25T14:57:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.25726", "title": "SIREN: Unified Multi-Granularity Semantic Interaction for Multi-Modal Lifelong User Interest Modeling", "published": "2026-05-25T11:33:27Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-tencent", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.25582", "title": "Extreme Region Policy Distillation", "published": "2026-05-25T08:32:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.25514", "title": "From Item-Only to Query-Item: Query-Conditioned Generative Search with QGS in Quark", "published": "2026-05-25T07:18:51Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "production-evidence", "search-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B03", "plan_reason": "6–5 月序列建模、生成搜索与多场景排序 — completed" }, { "arxiv_id": "2605.25360", "title": "Learning to Route Languages for Multilingual Policy Optimization", "published": "2026-05-25T02:28:41Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.25342", "title": "MATO: Multi-objective Personalized Alignment with Test-time Optimization for Large Language Models", "published": "2026-05-25T01:57:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.25338", "title": "CausalFlow: Causal Attribution and Counterfactual Repair for LLM Agent Failures", "published": "2026-05-25T01:47:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.25267", "title": "Latent Q-Barrier Shielding for Safe In-Context Reinforcement Learning", "published": "2026-05-24T21:45:28Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.25198", "title": "Hide to Guide: Learning via Semantic Masking", "published": "2026-05-24T17:59:06Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27450", "title": "Context Features Are Cheap: Rank-Aware Decomposition for Efficient Feature Interaction in Recommender Systems", "published": "2026-05-24T13:35:36Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking", "recommendation-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.25036", "title": "Language Bias in LVLMs: From In-Depth Analysis to Simple and Effective Mitigation", "published": "2026-05-24T12:23:13Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.25007", "title": "Meta-Modal Agent: Sequential Evidence Routing for Missing-Modality Candidate Reranking", "published": "2026-05-24T11:22:03Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.24960", "title": "Investigating the Interplay between Contextual and Parametric Chain-of-Thought Faithfulness under Optimization", "published": "2026-05-24T09:16:55Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.24949", "title": "APT-Agent: Automated Penetration Testing using Large Language Models", "published": "2026-05-24T08:54:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.24885", "title": "DTO: a Differentiable Training Objective for Effective Counterfactual Story Rewriting", "published": "2026-05-24T05:58:11Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.24597", "title": "Learning to Reason Efficiently with A* Post-Training", "published": "2026-05-23T14:28:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.24366", "title": "Structure-Aware RAG: Structured Retrieval Augmented Generation from Noisy Data for Conversational Agents", "published": "2026-05-23T03:07:33Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.24266", "title": "An Interactive Paradigm for Deep Research", "published": "2026-05-22T22:37:44Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00083", "title": "From Demonstrations to Rewards: Test-Time Prompt Optimization for VLM Reward Models", "published": "2026-05-22T16:04:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.23702", "title": "TubiFM: Unified Item, Carousel, and Search Ranking for Streaming Discovery", "published": "2026-05-22T14:53:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence", "search-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B03", "plan_reason": "6–5 月序列建模、生成搜索与多场景排序 — completed" }, { "arxiv_id": "2605.23657", "title": "OpenSkillEval: Automatically Auditing the Open Skill Ecosystem for LLM Agents", "published": "2026-05-22T14:09:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.23491", "title": "CoSPlay: Cooperative Self-Play at Test-Time with Self-Generated Code and Unit Test", "published": "2026-05-22T10:53:17Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.23384", "title": "Metacognition as Reward: Reinforcing LLM Reasoning via Knowledge and Regulation Signals", "published": "2026-05-22T08:54:37Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.23382", "title": "From Correctness to Preference: A Framework for Personalized Agentic Reinforcement Learning", "published": "2026-05-22T08:50:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.23244", "title": "Convex Optimization for Alignment and Preference Learning on a Single GPU", "published": "2026-05-22T05:25:00Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.23191", "title": "Expand More, Shrink Less: Shaping Effective-Rank Dynamics for Dense Scaling in Recommendation", "published": "2026-05-22T03:17:29Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.23071", "title": "The Efficiency Frontier: A Unified Framework for Cost-Performance Optimization in LLM Context Management", "published": "2026-05-21T22:03:25Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.23024", "title": "The Deterministic Horizon: Impossibility Results as Design Specifications for Trustworthy AI Systems", "published": "2026-05-21T20:48:35Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.22731", "title": "Post-Training is About States, Not Tokens: A State Distribution View of SFT, RL, and On-Policy Distillation", "published": "2026-05-21T17:03:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.22675", "title": "Self-Policy Distillation via Capability-Selective Subspace Projection", "published": "2026-05-21T16:18:41Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.22894", "title": "SCRIPT: Scalable Diffusion Policy with Multi-stage Training for Language-driven Physics-based Humanoid Control", "published": "2026-05-21T14:17:21Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.22511", "title": "Search-E1: Self-Distillation Drives Self-Evolution in Search-Augmented Reasoning", "published": "2026-05-21T14:00:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.22411", "title": "DeferMem: Query-Time Evidence Distillation via Reinforcement Learning for Long-Term Memory QA", "published": "2026-05-21T12:36:46Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.22219", "title": "SGR-Bench: Benchmarking Search Agents on State-Gated Retrieval", "published": "2026-05-21T09:22:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.22138", "title": "Efficient Agentic Reasoning Through Self-Regulated Simulative Planning", "published": "2026-05-21T08:11:54Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.22073", "title": "BRIDGE: Behavior-Guided Residual Integration with Dual-Frequency Graph Evidence", "published": "2026-05-21T07:10:40Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.22057", "title": "FlyRoute: Self-Evolving Agent Profiling via Data Flywheel for Adaptive Task Routing", "published": "2026-05-21T06:46:21Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.21965", "title": "SpecHop: Continuous Speculation for Accelerating Multi-Hop Retrieval Agents", "published": "2026-05-21T03:55:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.21883", "title": "Token-weighted Direct Preference Optimization with Attention", "published": "2026-05-21T01:43:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.21752", "title": "PEARL: Unbiased Percentile Estimation via Contrastive Learning for Industrial-Scale Livestream Recommendation", "published": "2026-05-20T21:25:41Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "industrial-ranking", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B03", "plan_reason": "6–5 月序列建模、生成搜索与多场景排序 — completed" }, { "arxiv_id": "2605.21482", "title": "DeepWeb-Bench: A Deep Research Benchmark Demanding Massive Cross-Source Evidence and Long-Horizon Derivation", "published": "2026-05-20T17:59:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.21463", "title": "Mem-$π$: Adaptive Memory through Learning When and What to Generate", "published": "2026-05-20T17:51:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.21295", "title": "TimeSRL: Generalizable Time-Series Behavioral Modeling via Semantic RL-Tuned LLMs -- A Case Study in Mental Health", "published": "2026-05-20T15:25:46Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.21266", "title": "How Much Online RL is Enough? Informative Rollouts for Offline Preference Optimization in RLVR", "published": "2026-05-20T14:53:43Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.21225", "title": "PREFINE: Preference-Based Implicit Reward and Cost Fine-Tuning for Safety Alignment", "published": "2026-05-20T14:19:45Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.21139", "title": "Distill to Think, Foresee to Act: Cognitive-Physical Reinforcement Learning for Autonomous Driving", "published": "2026-05-20T13:14:28Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.21063", "title": "APM: Evaluating Style Personalization in LLMs with Arbitrary Preference Mappings", "published": "2026-05-20T11:47:40Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "preference-optimization" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.20911", "title": "For How Long Should We Be Punching? Learning Action Duration in Fighting Games", "published": "2026-05-20T08:56:50Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.20833", "title": "MemGym: a Long-Horizon Memory Environment for LLM Agents", "published": "2026-05-20T07:25:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.20654", "title": "REFLECTOR: Internalizing Step-wise Reflection against Indirect Jailbreak", "published": "2026-05-20T03:16:15Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.20643", "title": "AVSD: Adaptive-View Self-Distillation by Balancing Consensus and Teacher-Specific Privileged Signals", "published": "2026-05-20T03:06:36Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.20559", "title": "Group-Aware Matrix Estimation and Latent Subspace Recovery", "published": "2026-05-19T23:22:32Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.20506", "title": "Reinforcing Human Behavior Simulation via Verbal Feedback", "published": "2026-05-19T21:23:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.28638", "title": "Learning Stateful Predictive Knowledge From Experience", "published": "2026-05-19T17:09:32Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "on-policy-distillation" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.20061", "title": "Rewarding Beliefs, Not Actions: Consistency-Guided Credit Assignment for Long-Horizon Agents", "published": "2026-05-19T16:19:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.19926", "title": "JAXenstein: Accelerated Benchmarking for First-Person Environments", "published": "2026-05-19T14:47:55Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.24005", "title": "LC-ERD: Mining Latent Logic for Self-Evolving Reasoning via Consistency-Regulated Reward Decomposition", "published": "2026-05-19T07:27:50Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.19444", "title": "Detecting and Mitigating the Correct-Answer Extinction Window in Test-Time Reinforcement Learning with Majority Voting", "published": "2026-05-19T06:58:44Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.19149", "title": "Agent Meltdowns: The Road to Hell Is Paved with Helpful Agents", "published": "2026-05-18T22:03:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.19130", "title": "EgoBabyVLM: Benchmarking Cross-Modal Learning from Naturalistic Egocentric Video Data", "published": "2026-05-18T21:30:54Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.18696", "title": "Ensembling Tabular Foundation Models - A Diversity Ceiling And A Calibration Trap", "published": "2026-05-18T17:32:57Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.20258", "title": "It Takes Two: Complementary Self-Distillation for Contextual Integrity in LLMs", "published": "2026-05-18T13:57:44Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.18374", "title": "Beyond Inference-Time Search: Reinforcement Learning Synthesizes Reusable Solvers", "published": "2026-05-18T13:21:40Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.18299", "title": "SD-Search: On-Policy Hindsight Self-Distillation for Search-Augmented Reasoning", "published": "2026-05-18T12:18:47Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.18261", "title": "Knowledge-to-Verification: Exploring RLVR for LLMs in Knowledge-Intensive Domains", "published": "2026-05-18T11:59:31Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.18133", "title": "An Empirical Study of Privacy Leakage Chains via Prompt Injection in Black-Box Chatbot Environments", "published": "2026-05-18T09:38:18Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.18044", "title": "Modality-Aware Identity Construction and Counterfactual Structure Learning for ID-Free Multimodal Recommendation", "published": "2026-05-18T08:35:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.17958", "title": "Enhancing the Code Reasoning Capabilities of LLMs via Consistency-based Reinforcement Learning", "published": "2026-05-18T07:12:44Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.17946", "title": "SVFSearch: A Multimodal Knowledge-Intensive Benchmark for Short-Video Frame Search in the Gaming Vertical Domain", "published": "2026-05-18T07:03:48Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.18918", "title": "ESLD (External Surrogate Latent Defense): A Latent-Space Architecture for Faster, Stronger Prompt-Injection Defense", "published": "2026-05-18T06:57:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.17863", "title": "DADF: A Distribution-Aware Debiasing Framework for Watch-Time Regression in Recommender Systems", "published": "2026-05-18T05:14:50Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "industrial-ranking", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B03", "plan_reason": "6–5 月序列建模、生成搜索与多场景排序 — completed" }, { "arxiv_id": "2605.17497", "title": "Self-Supervised On-Policy Distillation for Reasoning Language Models", "published": "2026-05-17T15:14:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.18899", "title": "Don't Let Bandit Feedback Pull Continual LLM-Recommender Updates Off Target", "published": "2026-05-17T11:10:44Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.17314", "title": "Weak-to-Strong Elicitation via Mismatched Wrong Drafts", "published": "2026-05-17T08:12:51Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.17295", "title": "DISA: Offline Importance Sampling for Distribution-Matching LLM-RL", "published": "2026-05-17T07:14:44Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.17036", "title": "Reliability and Effectiveness of Autonomous AI Agents in Supply Chain Management", "published": "2026-05-16T15:11:35Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.16999", "title": "Ranking-Aware Calibration for Reliable Multimodal Reinforcement Learning", "published": "2026-05-16T13:51:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.16479", "title": "Policy-Grounded Dynamic Facet Suggestions for Job Search", "published": "2026-05-15T17:30:45Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "search-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.16205", "title": "Context, Reasoning, and Hierarchy: A Cost-Performance Study of Compound LLM Agent Design in an Adversarial POMDP", "published": "2026-05-15T17:23:08Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.15963", "title": "PAGER: Bridging the Semantic-Execution Gap in Point-Precise Geometric GUI Control", "published": "2026-05-15T13:55:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.20588", "title": "AInterviewer: A Platform for Designing and Conducting AI-led Qualitative Interviews", "published": "2026-05-15T10:21:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.15777", "title": "SaaS-Bench: Can Computer-Use Agents Leverage Real-World SaaS to Solve Professional Workflows?", "published": "2026-05-15T09:35:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.15604", "title": "VSPO: Vector-Steered Policy Optimization for Behavioral Control", "published": "2026-05-15T04:31:06Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.16452", "title": "Peak-Detector: Explainable Peak Detection via Instruction-Tuned Large Language Models in Physiological Sign", "published": "2026-05-15T04:08:00Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.15529", "title": "Process Rewards with Learned Reliability", "published": "2026-05-15T01:57:11Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.15102", "title": "Improving Multi-turn Dialogue Consistency with Self-Recall Thinking", "published": "2026-05-14T17:20:14Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.15055", "title": "DiffusionOPD: A Unified Perspective of On-Policy Distillation in Diffusion Models", "published": "2026-05-14T16:49:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.15040", "title": "Orchard: An Open-Source Agentic Modeling Framework", "published": "2026-05-14T16:35:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd", "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.14929", "title": "A Hardware-Aware, Per-Layer Methodology for Post-Training Quantization of Large Language Models", "published": "2026-05-14T15:03:58Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.14853", "title": "Discrimination Is Generation: Unifying Ranking and Retrieval from a Tokenizer Perspective", "published": "2026-05-14T13:59:29Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking", "recommendation-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.14323", "title": "Dynamic Latent Routing", "published": "2026-05-14T03:35:46Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.14290", "title": "Web Agents Should Adopt the Plan-Then-Execute Paradigm", "published": "2026-05-14T02:48:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.14215", "title": "GenCircuit-RL: Reinforcement Learning from Hierarchical Verification for Genetic Circuit Design", "published": "2026-05-14T00:18:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.14071", "title": "Distribution Corrected Offline Data Distillation for Large Language Models", "published": "2026-05-13T19:47:31Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.13706", "title": "Identifying AI Web Scrapers Using Canary Tokens", "published": "2026-05-13T15:53:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.13918", "title": "CA2: Code-Aware Agent for Automated Game Testing", "published": "2026-05-13T12:52:35Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.18851", "title": "STRIDE: Learnable Stepwise Language Feedback for LLM Reasoning", "published": "2026-05-13T11:04:31Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.13110", "title": "A Multi-Agent Orchestration Framework for Venture Capital Due Diligence", "published": "2026-05-13T07:20:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.13052", "title": "RAG-Enhanced Large Language Models for Dynamic Content Expiration Prediction in Web Search", "published": "2026-05-13T06:20:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "search-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.12995", "title": "F-GRPO: Factorized Group-Relative Policy Optimization for Unified Candidate Generation and Ranking", "published": "2026-05-13T04:52:33Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.15220", "title": "Always Learning, Always Mixing: Efficient and Simple Data Mixing All The Time", "published": "2026-05-13T02:29:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.12645", "title": "Training LLMs with Reinforcement Learning for Intent-Aware Personalized Question Answering", "published": "2026-05-12T18:38:05Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.12227", "title": "A Recipe for Long-Context Reasoning in Large Language Models via On-Policy Optimization and Distillation", "published": "2026-05-12T15:04:18Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.18813", "title": "Composition of Memory Experts for Diffusion World Models", "published": "2026-05-12T09:43:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.11732", "title": "AgentDisCo: Towards Disentanglement and Collaboration in Open-ended Deep Research Agents", "published": "2026-05-12T08:14:15Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.11553", "title": "TwiSTAR:Think Fast, Think Slow, Then Act,Generative Recommendation with Adaptive Reasoning", "published": "2026-05-12T05:35:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recommendation-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.11433", "title": "FedMM: Federated Collaborative Signal Quantization for Multi-Market CTR Prediction", "published": "2026-05-12T02:32:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon", "priority-org-netflix" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.11328", "title": "Epistemic Uncertainty for Test-Time Discovery", "published": "2026-05-11T23:26:30Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.11299", "title": "Primal Generation, Dual Judgment: Self-Training from Test-Time Scaling", "published": "2026-05-11T22:34:45Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.11118", "title": "A Cascaded Generative Approach for e-Commerce Recommendations", "published": "2026-05-11T18:27:08Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "recommendation-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.10488", "title": "DeepRefine: Agent-Compiled Knowledge Refinement via Reinforcement Learning", "published": "2026-05-11T12:48:31Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.10377", "title": "PC3D: Zero-Shot Cooperation Across Variable Rosters via Personalized Context Distillation", "published": "2026-05-11T11:20:17Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.10367", "title": "AgentGR: Semantic-aware Agentic Group Decision-Making Simulator for Group Recommendation", "published": "2026-05-11T11:10:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.10194", "title": "TRACE: Distilling Where It Matters via Token-Routed Self On-Policy Alignment", "published": "2026-05-11T08:45:03Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.10158", "title": "Unsupervised Process Reward Models", "published": "2026-05-11T08:05:27Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.10090", "title": "CCD-Level and Load-Aware Thread Orchestration for In-Memory Vector ANNS on Multi-Core CPUs", "published": "2026-05-11T07:09:29Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.10064", "title": "MAGE: Multi-Agent Self-Evolution with Co-Evolutionary Knowledge Graphs", "published": "2026-05-11T06:39:51Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.16379", "title": "An Information-Theoretic Criterion for Efficient Data Synthesis", "published": "2026-05-11T01:27:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.09853", "title": "Exploration-Driven Optimization for Test-Time Large Language Model Reasoning", "published": "2026-05-11T01:10:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.09808", "title": "Quantifying the Utility of User Simulators for Building Collaborative LLM Assistants", "published": "2026-05-10T23:06:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.09638", "title": "Plan2Cleanse: Test-Time Backdoor Defense via Monte-Carlo Planning in Deep Reinforcement Learning", "published": "2026-05-10T16:34:16Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.09584", "title": "CLR-voyance: Reinforcing Open-Ended Reasoning for Inpatient Clinical Decision Support with Outcome-Aware Rubrics", "published": "2026-05-10T14:51:31Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.11017", "title": "Simpson's Paradox in Behavioral Curves: How Aggregation Distorts Parametric Models of User Dynamics", "published": "2026-05-10T14:44:13Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.09548", "title": "Crosslingual On-Policy Self-Distillation for Multilingual Reasoning", "published": "2026-05-10T14:06:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.09497", "title": "Don't Click That: Teaching Web Agents to Resist Deceptive Interfaces", "published": "2026-05-10T12:11:52Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.09359", "title": "Skill-R1: Agent Skill Evolution via Reinforcement Learning", "published": "2026-05-10T06:19:15Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.09321", "title": "OpenIIR: An Open Simulation Platform for Information Retrieval Research", "published": "2026-05-10T04:46:58Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recommendation-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.09283", "title": "A Prompt-Aware Structuring Framework for Reliable Reuse of AI-Generated Content in the Agentic Web", "published": "2026-05-10T03:16:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.09269", "title": "DeltaRubric: Generative Multimodal Reward Modeling via Joint Planning and Verification", "published": "2026-05-10T02:32:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.09253", "title": "Cornerstones or Stumbling Blocks? Deciphering the Rock Tokens in On-Policy Distillation", "published": "2026-05-10T01:41:43Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.16363", "title": "ORACLE: Anticipating Scams from Partial Trajectories in Streaming App Usage", "published": "2026-05-09T16:26:16Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.08828", "title": "When Agents Overtrust Environmental Evidence: An Extensible Agentic Framework for Benchmarking Evidence-Grounding Defects in LLM Agents", "published": "2026-05-09T09:32:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.08731", "title": "Choosing a JPEG Decoder for PyTorch DataLoaders: Workload-Specific Throughput on Four CPUs", "published": "2026-05-09T06:34:17Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.08703", "title": "RewardHarness: Self-Evolving Agentic Post-Training", "published": "2026-05-09T05:32:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.08583", "title": "Source or It Didn't Happen: A Multi-Agent Framework for Citation Hallucination Detection", "published": "2026-05-09T00:53:24Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07817", "title": "GazeVLM: Active Vision via Internal Attention Control for Multimodal Reasoning", "published": "2026-05-08T14:49:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07775", "title": "POETS: Uncertainty-Aware LLM Optimization via Compute-Efficient Policy Ensembles", "published": "2026-05-08T14:16:32Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07701", "title": "Guidance Is Not a Hyperparameter: Learning Dynamic Control in Diffusion Language Models", "published": "2026-05-08T13:12:32Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07637", "title": "Learning to Communicate Locally for Large-Scale Multi-Agent Pathfinding", "published": "2026-05-08T12:05:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.14115", "title": "DialogueVPR: Towards Conversational Visual Place Recognition", "published": "2026-05-08T12:04:43Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07579", "title": "Your Language Model is Its Own Critic: Reinforcement Learning with Value Estimation from Actor's Internal States", "published": "2026-05-08T10:49:36Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07510", "title": "InterLV-Search: Benchmarking Interleaved Multimodal Agentic Search", "published": "2026-05-08T09:41:07Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07465", "title": "SEIF: Self-Evolving Reinforcement Learning for Instruction Following", "published": "2026-05-08T09:13:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07461", "title": "Think-with-Rubrics: From External Evaluator to Internal Reasoning Guidance", "published": "2026-05-08T09:08:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.08283", "title": "HTPO: Towards Exploration-Exploitation Balanced Policy Optimization via Hierarchical Token-level Objective Control", "published": "2026-05-08T07:38:35Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07153", "title": "Beyond Reasoning: Reinforcement Learning Unlocks Parametric Knowledge in LLMs", "published": "2026-05-08T02:40:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07147", "title": "MathlibPR: Pull Request Merge-Readiness Benchmark for Formal Mathematical Libraries", "published": "2026-05-08T02:32:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07137", "title": "Adaptive Negative Reinforcement for LLM Reasoning:Dynamically Balancing Correction and Diversity in RLVR", "published": "2026-05-08T02:13:33Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07134", "title": "Region4Web: Rethinking Observation Space Granularity for Web Agents", "published": "2026-05-08T02:11:23Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07129", "title": "RRCM: Ranking-Driven Retrieval over Collaborative and Meta Memories for LLM Recommendation", "published": "2026-05-08T02:07:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.07125", "title": "An Embarrassingly Simple Graph Heuristic Reveals Shortcut-Solvable Benchmarks for Sequential Recommendation", "published": "2026-05-08T02:00:11Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.16344", "title": "A Production-Ready RL Framework for Personalized Utility Tuning with Pareto Sweeping in Pinterest Recommender Systems", "published": "2026-05-08T01:48:49Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "priority-org-pinterest", "recommendation-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B04", "plan_reason": "推荐 RL、知识迁移与 Semantic ID — completed" }, { "arxiv_id": "2605.07076", "title": "Self-Consolidating Language Models: Continual Knowledge Incorporation from Context", "published": "2026-05-08T00:50:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07058", "title": "MedExAgent: Training LLM Agents to Ask, Examine, and Diagnose in Noisy Clinical Environments", "published": "2026-05-08T00:12:23Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.07039", "title": "PACEvolve++: Improving Test-time Learning for Evolutionary Search Agents", "published": "2026-05-07T23:38:50Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.06981", "title": "Bridging Textual Profiles and Latent User Embeddings for Personalization", "published": "2026-05-07T21:59:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-amazon", "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.06947", "title": "Rollback-Free Stable Brick Structures Generation", "published": "2026-05-07T21:06:44Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.06941", "title": "Causal-Aware Foundation-Model for Bilevel Optimization in Discrete Choice Settings", "published": "2026-05-07T20:55:47Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.06850", "title": "How to Compress KV Cache in RL Post-Training? Shadow Mask Distillation for Memory-Efficient Alignment", "published": "2026-05-07T18:51:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.06635", "title": "Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents", "published": "2026-05-07T17:46:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.06445", "title": "Constraint Decay: The Fragility of LLM Agents in Backend Code Generation", "published": "2026-05-07T15:44:40Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.06200", "title": "A$^2$TGPO: Agentic Turn-Group Policy Optimization with Adaptive Turn-level Clipping", "published": "2026-05-07T13:09:31Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.06040", "title": "Novelty-based Tree-of-Thought Search for LLM Reasoning and Planning", "published": "2026-05-07T11:28:53Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.06017", "title": "Matrix-Decoupled Concentration for Autoregressive Sequences: Dimension-Free Guarantees for Sparse Long-Context Rewards", "published": "2026-05-07T11:12:59Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.05855", "title": "Bridging Passive and Active: Enhancing Conversation Starter Recommendation via Active Expression Modeling", "published": "2026-05-07T08:26:24Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.05739", "title": "Multi-Dimensional Behavioral Evaluation of Agentic Stock Prediction Systems Using Large Language Model Judges with Closed-Loop Reinforcement Learning Feedback", "published": "2026-05-07T06:31:34Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.05730", "title": "Effective Knowledge Transfer for Multi-Task Recommendation Models", "published": "2026-05-07T06:22:20Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B04", "plan_reason": "推荐 RL、知识迁移与 Semantic ID — completed" }, { "arxiv_id": "2605.05096", "title": "CapsID: Soft-Routed Variable-Length Semantic IDs for Generative Recommendation", "published": "2026-05-06T16:33:13Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.04895", "title": "Regime-Conditioned Evaluation in Multi-Context Bayesian Optimization", "published": "2026-05-06T13:27:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "search-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.04719", "title": "Every Step Counts: Step-Level Credit Assignment for Tool-Integrated Text-to-SQL", "published": "2026-05-06T10:10:31Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.04470", "title": "CRAFT: Counterfactual-to-Interactive Reinforcement Fine-Tuning for Driving Policies", "published": "2026-05-06T03:49:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.04261", "title": "Laundering AI Authority with Adversarial Examples", "published": "2026-05-05T19:55:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.03952", "title": "MOSAIC-Bench: Measuring Compositional Vulnerability Induction in Coding Agents", "published": "2026-05-05T16:38:23Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.03788", "title": "Say the Mission, Execute the Swarm: Agent-Enhanced LLM Reasoning in the Web-of-Drones", "published": "2026-05-05T14:14:57Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.00049", "title": "Measuring and Mitigating Bias in Code Generated by Large Language Models", "published": "2026-05-05T14:03:16Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.06702", "title": "CASCADE: Case-Based Continual Adaptation for Large Language Models During Deployment", "published": "2026-05-05T12:16:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.08186", "title": "Rethinking Entropy Minimization in Test-Time Adaptation for Autoregressive Models", "published": "2026-05-05T12:00:06Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.03403", "title": "GRPO-TTA: Test-Time Visual Tuning for Vision-Language Models via GRPO-Driven Reinforcement Learning", "published": "2026-05-05T06:23:20Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.03242", "title": "Enhancing Agent Safety Judgment: Controlled Benchmark Rewriting and Analogical Reasoning for Deceptive Out-of-Distribution Scenarios", "published": "2026-05-05T00:21:00Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.03129", "title": "PIIGuard: Mitigating PII Harvesting under Adversarial Sanitization", "published": "2026-05-04T20:13:22Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.16315", "title": "A Structural Threshold in Decision Capacity Governs Collapse in Self-Play Reinforcement Learning", "published": "2026-05-04T19:09:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.02348", "title": "Decoding-Time Debiasing via Process Reward Models: From Controlled Fill-in to Open-Ended Generation", "published": "2026-05-04T08:51:34Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.02346", "title": "APIOT: Autonomous Vulnerability Management Across Bare-Metal Industrial OT Networks", "published": "2026-05-04T08:47:28Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.02168", "title": "Planner Matters! An Efficient and Unbalanced Multi-agent Collaboration Framework for Long-horizon Planning", "published": "2026-05-04T02:58:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.07546", "title": "Beyond Item IDs: Scaling Short-Form-Video Recommendation via Semantic-Native Long Sequence Modeling", "published": "2026-05-04T00:00:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.01771", "title": "The Compliance Gap: Why AI Systems Promise to Follow Process Instructions but Don't", "published": "2026-05-03T08:11:15Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.01567", "title": "Feedback-Normalized Developer Memory for Reinforcement-Learning Coding Agents: A Safety-Gated MCP Architecture", "published": "2026-05-02T18:37:36Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.01489", "title": "SciResearcher: Scaling Deep Research Agents for Frontier Scientific Reasoning", "published": "2026-05-02T15:26:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.00007", "title": "BaRA: Budget-constrained and Reliable Web Data Collection Agent", "published": "2026-05-02T08:09:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.01224", "title": "Lost in the Tower of Babel: The Adverse Effects of Incidental Multilingualism in LLMs", "published": "2026-05-02T03:39:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.01203", "title": "GR-Ben: A General Reasoning Benchmark for Evaluating Process Reward Models", "published": "2026-05-02T02:41:48Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.15206", "title": "AgentStop: Terminating Local AI Agents Early to Save Energy in Consumer Devices", "published": "2026-05-01T14:45:37Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.00702", "title": "Learning How and What to Memorize: Cognition-Inspired Two-Stage Optimization for Evolving Memory", "published": "2026-05-01T14:45:20Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.00553", "title": "Stable-GFlowNet: Toward Diverse and Robust LLM Red-Teaming via Contrastive Trajectory Balance", "published": "2026-05-01T10:42:08Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.00353", "title": "Negative Data Mining for Contrastive Learning in Dense Retrieval at IKEA.com", "published": "2026-05-01T02:32:02Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "search-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.00298", "title": "Data Deletion Can Help in Adaptive RL", "published": "2026-04-30T23:58:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.00237", "title": "Bayesian Optimization in Linear Time", "published": "2026-04-30T21:16:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.27840", "title": "CastFlow: Learning Role-Specialized Agentic Workflows for Time Series Forecasting", "published": "2026-04-30T13:24:42Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.00068", "title": "Human-in-the-Loop Meta Bayesian Optimization for Fusion Energy and Scientific Applications", "published": "2026-04-30T10:06:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.27472", "title": "PRTS: A Primitive Reasoning and Tasking System via Contrastive Representations", "published": "2026-04-30T06:14:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.27321", "title": "Toward Autonomous SOC Operations: End-to-End LLM Framework for Threat Detection, Query Generation, and Resolution in Security Operations", "published": "2026-04-30T02:06:46Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.27253", "title": "AutoSurfer -- Teaching Web Agents through Comprehensive Surfing, Learning, and Modeling", "published": "2026-04-29T22:57:35Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.27117", "title": "A Gated Hybrid Contrastive Collaborative Filtering Recommendation", "published": "2026-04-29T19:10:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.26573", "title": "PAINT: Partial-Solution Adaptive Interpolated Training for Self-Distilled Reasoners", "published": "2026-04-29T11:56:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.26516", "title": "Lyapunov-Guided Self-Alignment: Test-Time Adaptation for Offline Safe Reinforcement Learning", "published": "2026-04-29T10:32:18Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.21497", "title": "Autonomous LLM Agents & CTFs: A Second Look", "published": "2026-04-29T09:42:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.26390", "title": "Meta-Learning and Targeted Differential Privacy to Improve the Accuracy-Privacy Trade-off in Recommendations", "published": "2026-04-29T08:00:20Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.26173", "title": "Entropy Centroids as Intrinsic Rewards for Test-Time Scaling", "published": "2026-04-28T23:27:21Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.19354", "title": "Granularity-Regulated Adaptive Computational Efficiency for Optimal Verification in Test-Time Scaling", "published": "2026-04-28T18:19:16Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.25834", "title": "Action-Aware Generative Sequence Modeling for Short Video Recommendation", "published": "2026-04-28T16:41:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-kuaishou", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.25732", "title": "Personalized Multi-Interest Modeling for Cross-Domain Recommendation to Cold-Start Users", "published": "2026-04-28T15:01:27Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.25562", "title": "SnapGuard: Lightweight Prompt Injection Detection for Screenshot-Based Web Agents", "published": "2026-04-28T12:32:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.23939", "title": "DRIVE: Modeling Skills at the Reasoning and Interaction Levels for Web Agents under Continual Learning", "published": "2026-04-28T11:39:20Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.24964", "title": "Odysseys: Benchmarking Web Agents on Realistic Long Horizon Tasks", "published": "2026-04-27T20:05:41Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.24957", "title": "Compute Aligned Training: Optimizing for Test Time Inference", "published": "2026-04-27T19:52:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.24594", "title": "Skill Retrieval Augmentation for Agentic AI", "published": "2026-04-27T15:19:59Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.24472", "title": "Modeling Behavioral Intensity and Transitions for Generative Recommendation", "published": "2026-04-27T13:40:35Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-taobao" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.24357", "title": "DPRM: A Plug-in Doob h transform-induced Token-Ordering Module for Diffusion Language Models", "published": "2026-04-27T11:50:26Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.24198", "title": "Rewarding the Scientific Process: Process-Level Reward Modeling for Agentic Data Analysis", "published": "2026-04-27T09:00:30Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.24806", "title": "Versioned Late Materialization for Ultra-Long Sequence Training in Recommendation Systems at Scale", "published": "2026-04-27T06:41:39Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.24029", "title": "DeepTaxon: An Interpretable Retrieval-Augmented Multimodal Framework for Unified Species Identification and Discovery", "published": "2026-04-27T04:30:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.23837", "title": "One Size Fits None: Heuristic Collapse in LLM Investment Advice", "published": "2026-04-26T18:45:11Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.23522", "title": "Beyond Static Collision Handling: Adaptive Semantic ID Learning for Multimodal Recommendation at Industrial Scale", "published": "2026-04-26T03:59:35Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "priority-org-kuaishou", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B04", "plan_reason": "推荐 RL、知识迁移与 Semantic ID — completed" }, { "arxiv_id": "2604.23488", "title": "Do Prompt-Elicited Trajectories Reflect Training-Time Reward Hacking? A Systematic Study on Monitoring Training-Time Reward Hacking in Code Generation", "published": "2026-04-26T01:26:50Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.23333", "title": "Process Supervision of Confidence Margin for Calibrated LLM Reasoning", "published": "2026-04-25T14:40:13Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.23318", "title": "Hidden States Know Where Reasoning Diverges: Credit Assignment via Span-Level Wasserstein Distance", "published": "2026-04-25T14:11:23Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.23088", "title": "Code Broker: A Multi-Agent System for Automated Code Quality Assessment", "published": "2026-04-25T00:53:59Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.23022", "title": "CASP: Support-Aware Offline Policy Selection for Two-Stage Recommender Systems", "published": "2026-04-24T21:21:10Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recommendation-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.22558", "title": "SOLAR-RL: Semi-Online Long-horizon Assignment Reinforcement Learning", "published": "2026-04-24T13:53:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.22229", "title": "Preserve Support, Not Correspondence: Dynamic Routing for Offline Reinforcement Learning", "published": "2026-04-24T05:07:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.22191", "title": "Behavioral Canaries: Auditing Private Retrieved Context Usage in RL Fine-Tuning", "published": "2026-04-24T03:38:52Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward", "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2606.11209", "title": "ProcessThinker: Enhancing Multi-modal Large Language Models Reasoning via Rollout-based Process Reward", "published": "2026-04-23T21:25:47Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.22873", "title": "When Policies Cannot Be Retrained: A Unified Closed-Form View of Post-Training Steering in Offline Reinforcement Learning", "published": "2026-04-23T20:20:11Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.21327", "title": "Understanding and Mitigating Spurious Signal Amplification in Test-Time Reinforcement Learning for Math Reasoning", "published": "2026-04-23T06:32:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.20755", "title": "V-tableR1: Process-Supervised Multimodal Table Reasoning with Critic-Guided Policy Optimization", "published": "2026-04-22T16:44:33Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.20714", "title": "Learning to Evolve: A Self-Improving Framework for Multi-Agent Systems via Textual Parameter Graph Optimization", "published": "2026-04-22T16:00:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.20659", "title": "GRPO-VPS: Enhancing Group Relative Policy Optimization with Verifiable Process Supervision for Effective Reasoning", "published": "2026-04-22T15:08:58Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.20398", "title": "WebGen-R1: Incentivizing Large Language Models to Generate Functional and Aesthetic Websites with Reinforcement Learning", "published": "2026-04-22T10:04:46Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.20316", "title": "R2IF: Aligning Reasoning with Decisions via Composite Rewards for Interpretable LLM Function Calling", "published": "2026-04-22T08:13:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.20273", "title": "ActuBench: A Multi-Agent LLM Pipeline for Generation and Evaluation of Actuarial Reasoning Tasks", "published": "2026-04-22T07:20:03Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.20135", "title": "AFMRL: Attribute-Enhanced Fine-Grained Multi-Modal Representation Learning in E-commerce", "published": "2026-04-22T03:02:21Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.19859", "title": "DR-Venus: Towards Frontier Edge-Scale Deep Research Agents with Only 10K Open Data", "published": "2026-04-21T17:59:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.19730", "title": "FASTER: Value-Guided Sampling for Fast RL", "published": "2026-04-21T17:52:17Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.19857", "title": "Rethinking Reinforcement Fine-Tuning in LVLM: Convergence, Reward Decomposition, and Generalization", "published": "2026-04-21T17:21:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.19656", "title": "Pause or Fabricate? Training Language Models for Grounded Reasoning", "published": "2026-04-21T16:45:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.22840", "title": "AeSlides: Incentivizing Aesthetic Layout in LLM-Based Slide Generation via Verifiable Rewards", "published": "2026-04-21T11:59:03Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.19354", "title": "Do Agents Dream of Root Shells? Partial-Credit Evaluation of LLM Agents in Capture the Flag Challenges", "published": "2026-04-21T11:35:33Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.19128", "title": "GraphRAG-IRL: Personalized Recommendation with Graph-Grounded Inverse Reinforcement Learning and LLM Re-ranking", "published": "2026-04-21T06:22:29Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recommendation-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.19024", "title": "Policy Gradient Primal-Dual Method for Safe Reinforcement Learning from Human Feedback", "published": "2026-04-21T03:20:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.18789", "title": "ARES: Adaptive Red-Teaming and End-to-End Repair of Policy-Reward System", "published": "2026-04-20T19:54:47Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.18779", "title": "Mango: Multi-Agent Web Navigation via Global-View Optimization", "published": "2026-04-20T19:42:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.16302", "title": "Reducing Credit Assignment Variance via Counterfactual Reasoning Paths", "published": "2026-04-20T13:33:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.18224", "title": "WebCompass: Towards Multimodal Web Coding Evaluation for Code Language Models", "published": "2026-04-20T13:09:38Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.18187", "title": "Audio-DeepThinker: Progressive Reasoning-Aware Reinforcement Learning for High-Quality Chain-of-Thought Emergence in Audio Language Models", "published": "2026-04-20T12:43:00Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.18146", "title": "Modular Representation Compression: Adapting LLMs for Efficient and Effective Recommendations", "published": "2026-04-20T12:08:58Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.17982", "title": "Mitigating Multimodal Hallucination via Phase-wise Self-reward", "published": "2026-04-20T09:04:49Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.17957", "title": "Process Reward Models Meet Planning: Generating Precise and Scalable Datasets for Step-Level Rewards", "published": "2026-04-20T08:39:13Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.17912", "title": "Learning to Correct: Calibrated Reinforcement Learning for Multi-Attempt Chain-of-Thought", "published": "2026-04-20T07:42:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.17878", "title": "RankUp: Towards High-rank Representations for Large Scale Advertising Recommender Systems", "published": "2026-04-20T06:40:27Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.17821", "title": "WebUncertainty: Dual-Level Uncertainty Driven Planning and Reasoning For Autonomous Web Agent", "published": "2026-04-20T05:19:49Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.17456", "title": "TrafficClaw: A Generalizable LLM Agent in the Unified Physical Environment for Urban Traffic Control", "published": "2026-04-19T14:17:56Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.17406", "title": "EvoMaster: A Foundational Evolving Agent Framework for Agentic Science at Scale", "published": "2026-04-19T12:26:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.05226", "title": "Internalizing Outcome Supervision into Process Supervision: A New Paradigm for Reinforcement Learning for Reasoning", "published": "2026-04-19T10:33:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.17282", "title": "MedPRMBench: A Fine-grained Benchmark for Process Reward Models in Medical Reasoning", "published": "2026-04-19T06:44:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "process-reward" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.17259", "title": "HORIZON: A Benchmark for In-the-wild User Behaviour Modeling", "published": "2026-04-19T04:45:48Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.17091", "title": "GenericAgent: A Token-Efficient Self-Evolving LLM Agent via Contextual Information Density Maximization (V1.0)", "published": "2026-04-18T17:59:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.17020", "title": "Beyond Static Benchmarks: Synthesizing Harmful Content via Persona-based Simulation for Robust Evaluation", "published": "2026-04-18T14:58:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.16968", "title": "On Safety Risks in Experience-Driven Self-Evolving Agents", "published": "2026-04-18T11:32:36Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.16004", "title": "AgentV-RL: Scaling Reward Modeling with Agentic Verifier", "published": "2026-04-17T12:27:36Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.15937", "title": "Polarization by Default: Auditing Recommendation Bias in LLM-Based Content Curation", "published": "2026-04-17T10:55:21Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.15618", "title": "Majority Voting for Code Generation", "published": "2026-04-17T01:51:36Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.15577", "title": "Reward Weighted Classifier-Free Guidance as Policy Improvement in Autoregressive Models", "published": "2026-04-16T23:13:22Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.15149", "title": "LLMs Gaming Verifiers: RLVR can Lead to Reward Hacking", "published": "2026-04-16T15:30:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.14683", "title": "DR$^{3}$-Eval: Towards Realistic and Reproducible Deep Research Evaluation", "published": "2026-04-16T06:40:02Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.14448", "title": "MARCA: A Checklist-Based Benchmark for Multilingual Web Search", "published": "2026-04-15T21:54:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.14352", "title": "PROXIMA: A Reliability Scoring Framework for Proxy Metrics in Online Controlled Experiments", "published": "2026-04-15T19:10:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.14265", "title": "Reinforcement Learning via Value Gradient Flow", "published": "2026-04-15T17:12:56Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.14032", "title": "Hierarchical Reinforcement Learning with Runtime Safety Shielding for Power Grid Operation", "published": "2026-04-15T16:11:10Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.20209", "title": "NaP-Control: Navigating Diffusion Prior for Versatile and Fast Character Control", "published": "2026-04-15T14:51:32Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.13796", "title": "Driving Engagement in Daily Fantasy Sports with a Scalable and Urgency-Aware Ranking Engine", "published": "2026-04-15T12:34:20Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.13737", "title": "TokenFormer: Unify the Multi-Field and Sequential Recommendation Worlds", "published": "2026-04-15T11:25:46Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-tencent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.13733", "title": "Vision-Language-Action Jump-Starting for Reinforcement Learning Robotic Agents", "published": "2026-04-15T11:17:54Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.13699", "title": "MIND: AI Co-Scientist for Material Research", "published": "2026-04-15T10:27:01Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2607.18240", "title": "Calibrated Selective Fact-Checking via Evidence Chain Evaluation", "published": "2026-04-15T02:48:05Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.13418", "title": "MERRIN: A Benchmark for Multimodal Evidence Retrieval and Reasoning in Noisy Web Environments", "published": "2026-04-15T02:37:47Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.13318", "title": "WebXSkill: Skill Learning for Autonomous Web Agents", "published": "2026-04-14T21:48:15Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.12965", "title": "Efficient Retrieval Scaling with Hierarchical Indexing for Large Scale Recommendation", "published": "2026-04-14T16:59:03Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.12666", "title": "From Imitation to Discrimination: Progressive Curriculum Learning for Robust Web Navigation", "published": "2026-04-14T12:37:45Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.14223", "title": "TRACE: A Conversational Framework for Sustainable Tourism Recommendation with Agentic Counterfactual Explanations", "published": "2026-04-14T12:35:21Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.04076", "title": "A Regulatory Governance Framework for AI-Driven Financial Fraud Detection in U.S. Banking: Integrating OCC, SR 11-7, CFPB, and FinCEN Compliance Requirements for Model Development, Validation, and Monitoring Lifecycles", "published": "2026-04-14T08:30:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.12234", "title": "UniRec: Bridging the Expressive Gap between Generative and Discriminative Recommendation via Chain-of-Attribute", "published": "2026-04-14T03:13:50Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence", "search-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B04", "plan_reason": "推荐 RL、知识迁移与 Semantic ID — completed" }, { "arxiv_id": "2604.11790", "title": "ClawGuard: A Runtime Security Framework for Tool-Augmented LLM Agents Against Indirect Prompt Injection", "published": "2026-04-13T17:55:11Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.11626", "title": "RationalRewards: Reasoning Rewards Scale Visual Generation Both Training and Test Time", "published": "2026-04-13T15:38:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.11440", "title": "R3-VAE: Reference Vector-Guided Rating Residual Quantization VAE for Generative Recommendation", "published": "2026-04-13T13:23:29Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2606.00015", "title": "SortingHat: Redefining Operating Systems Education with a Tailored Digital Teaching Assistant", "published": "2026-04-13T13:17:54Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.10911", "title": "EvoNash-MARL: A Closed-Loop Multi-Agent Reinforcement Learning Framework for Medium-Horizon Equity Allocation", "published": "2026-04-13T02:24:32Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.10471", "title": "SID-Coord: Coordinating Semantic IDs for ID-based Ranking in Short-Video Search", "published": "2026-04-12T05:51:35Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "search-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.00846", "title": "ClinicBot: A Guideline-Grounded Clinical Chatbot with Prioritized Evidence RAG and Verifiable Citations", "published": "2026-04-11T00:37:12Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.08523", "title": "ClawBench: Can AI Agents Complete Everyday Online Tasks?", "published": "2026-04-09T17:57:13Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.08468", "title": "TTVS: Boosting Self-Exploring Reinforcement Learning via Test-time Variational Synthesis", "published": "2026-04-09T17:03:49Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.08181", "title": "Long-Term Embeddings for Balanced Personalization", "published": "2026-04-09T12:36:29Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.07941", "title": "Large Language Model Post-Training: A Unified View of Off-Policy and On-Policy Learning", "published": "2026-04-09T08:00:37Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.07720", "title": "Towards Knowledgeable Deep Research: Framework and Benchmark", "published": "2026-04-09T02:06:27Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.06940", "title": "A First Guess is Rarely the Final Answer: Learning to Search in the Traveling Salesperson Problem", "published": "2026-04-08T11:04:45Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.06928", "title": "Leveraging LLMs and Heterogeneous Knowledge Graphs for Persona-Driven Session-Based Recommendation", "published": "2026-04-08T10:40:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.27374", "title": "ICG: Improving Cover Image Generation via MLLM-based Prompting and Personalized Preference Alignment", "published": "2026-04-08T06:36:54Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.06718", "title": "CASE: Cadence-Aware Set Encoding for Large-Scale Next Basket Repurchase Recommendation", "published": "2026-04-08T06:31:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.06474", "title": "DataSTORM: Deep Research on Large-Scale Databases using Exploratory Data Analysis and Data Storytelling", "published": "2026-04-07T21:19:26Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.06367", "title": "WebSP-Eval: Evaluating Web Agents on Website Security and Privacy Tasks", "published": "2026-04-07T18:43:21Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.05424", "title": "PRISM-MCTS: Learning from Reasoning Trajectories with Metacognitive Reflection", "published": "2026-04-07T04:37:35Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.05329", "title": "Semantic Trimming and Auxiliary Multi-step Prediction for Generative Recommendation", "published": "2026-04-07T02:00:23Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.09698", "title": "Evaluating Scene-based In-Situ Item Labeling for Immersive Conversational Recommendation", "published": "2026-04-06T23:52:35Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2605.21491", "title": "Teaching Language Models to Forecast Research Success Through Comparative Idea Evaluation", "published": "2026-04-06T19:03:11Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.04923", "title": "Stratifying Reinforcement Learning with Signal Temporal Logic", "published": "2026-04-06T17:58:58Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.04898", "title": "QED-Nano: Teaching a Tiny Model to Prove Hard Theorems", "published": "2026-04-06T17:44:25Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.04872", "title": "Synthetic Sandbox for Training Machine Learning Engineering Agents", "published": "2026-04-06T17:19:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.04808", "title": "Selecting Decision-Relevant Concepts in Reinforcement Learning", "published": "2026-04-06T16:12:42Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.04767", "title": "Cog-DRIFT: Exploration on Adaptively Reformulated Instances Enables Learning from Hard Reasoning Problems", "published": "2026-04-06T15:38:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.04749", "title": "AI Trust OS -- A Continuous Governance Framework for Autonomous AI Observability and Zero-Trust Compliance in Enterprise Environments", "published": "2026-04-06T15:14:10Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.04263", "title": "Commercial Persuasion in AI-Mediated Conversations", "published": "2026-04-05T20:51:55Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.03949", "title": "Semantic IDs for Recommender Systems at Snapchat: Use Cases, Technical Challenges, and Design Choices", "published": "2026-04-05T03:46:52Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.04976", "title": "Tencent Advertising Algorithm Challenge 2025: All-Modality Generative Recommendation", "published": "2026-04-04T17:05:15Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking", "priority-org-tencent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.03724", "title": "Rank, Don't Generate: Statement-level Ranking for Explainable Recommendation", "published": "2026-04-04T13:01:45Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.03523", "title": "Optimizing Neurorobot Policy under Limited Demonstration Data through Preference Regret", "published": "2026-04-04T00:03:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.03465", "title": "The Tool Illusion: Rethinking Tool Use in Web Agents", "published": "2026-04-03T21:18:26Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.03159", "title": "BibTeX Citation Errors in Scientific Publishing Agents: Evaluation and Mitigation", "published": "2026-04-03T16:30:58Z", "tracks": [ "agent" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "web-agent" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.01476", "title": "From Rebound to Remedy: Understanding and Mitigating Reward Hacking via Representation Engineering", "published": "2026-04-01T23:33:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.09671", "title": "Belief-State RWKV for Reinforcement Learning under Partial Observability", "published": "2026-04-01T22:28:38Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.01430", "title": "Improving Latent Generalization Using Test-time Compute", "published": "2026-04-01T22:10:21Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.01302", "title": "Scaling Reasoning Tokens via RL and Parallel Thinking: Evidence From Competitive Programming", "published": "2026-04-01T18:05:49Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.01142", "title": "Deep Reinforcement Learning for Robotic Manipulation under Distribution Shift with Bounded Extremum Seeking", "published": "2026-04-01T16:59:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.00698", "title": "Learning to Hint for Reinforcement Learning", "published": "2026-04-01T09:58:08Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.00438", "title": "TR-ICRL: Test-Time Rethinking for In-Context Reinforcement Learning", "published": "2026-04-01T03:34:05Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.29259", "title": "Aligning Multimodal Sequential Recommendations via Robust Direct Preference Optimization with Sparse MoE", "published": "2026-03-31T04:49:32Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.28200", "title": "A Deep Reinforcement Learning Framework for Closed-loop Guidance of Fish Schools via Virtual Agents", "published": "2026-03-30T09:10:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.28124", "title": "RCLRec: Reverse Curriculum Learning for Modeling Sparse Conversions in Generative Recommendation", "published": "2026-03-30T07:41:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.27522", "title": "Hidden Ads: Behavior Triggered Semantic Backdoors for Advertisement Injection in Vision Language Models", "published": "2026-03-29T05:14:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.27389", "title": "Prediction-Based Markov Violation Scores for Detecting Non-Markovian Observations in Reinforcement Learning", "published": "2026-03-28T19:42:27Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.26100", "title": "Rethinking Recommendation Paradigms: From Pipelines to Agentic Recommender Systems", "published": "2026-03-27T06:14:18Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.26085", "title": "AgenticRS-Architecture: System Design for Agentic Recommender Systems", "published": "2026-03-27T05:26:31Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.25464", "title": "Maximum Entropy Behavior Exploration for Sim2Real Zero-Shot Reinforcement Learning", "published": "2026-03-26T14:07:01Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.25070", "title": "An Explainable Ensemble Learning Framework for Crop Classification with Optimized Feature Pyramids and Deep Networks", "published": "2026-03-26T06:13:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.24963", "title": "Design Once, Deploy at Scale: Template-Driven ML Development for Large Model Ecosystems", "published": "2026-03-26T02:58:26Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "deployment-evidence", "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.24840", "title": "Prune as You Generate: Online Rollout Pruning for Faster and Better RLVR", "published": "2026-03-25T22:10:36Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.24226", "title": "UniScale: Synergistic Entire Space Data and Model Scaling for Search Ranking", "published": "2026-03-25T12:00:26Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "advertising-ranking", "industrial-ranking", "production-evidence", "search-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B04", "plan_reason": "推荐 RL、知识迁移与 Semantic ID — completed" }, { "arxiv_id": "2604.16379", "title": "LLMAR: A Tuning-Free Recommendation Framework for Sparse and Text-Rich Industrial Domains", "published": "2026-03-25T02:49:27Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.22916", "title": "GateSID: Adaptive Gating for Balancing Semantic and Collaborative Signals in Recommendation", "published": "2026-03-24T08:04:41Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B04", "plan_reason": "推荐 RL、知识迁移与 Semantic ID — completed" }, { "arxiv_id": "2603.22117", "title": "On the Direction of RLVR Updates for LLM Reasoning: Identification and Exploitation", "published": "2026-03-23T15:42:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.21743", "title": "CellFluxRL: Biologically-Constrained Virtual Cell Modeling via Reinforcement Learning", "published": "2026-03-23T09:33:18Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2605.20189", "title": "SOLAR: A Self-Optimizing Open-Ended Autonomous Agent for Lifelong Learning and Continual Adaptation", "published": "2026-03-23T07:18:02Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.21481", "title": "TagLLM: A Fine-Grained Tag Generation Approach for Note Recommendation", "published": "2026-03-23T02:01:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.21465", "title": "DRTriton: Large-Scale Synthetic Data Driven Reinforcement Learning for Triton Kernel Generation", "published": "2026-03-23T00:59:35Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.21354", "title": "The Workload-Router-Pool Architecture for LLM Inference Optimization: A Vision Paper from the vLLM Semantic Router Project", "published": "2026-03-22T18:30:11Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.20926", "title": "Deep Adaptive Rate Allocation in Volatile Heterogeneous Wireless Networks", "published": "2026-03-21T20:03:56Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.20723", "title": "Algorithmic Audit of Personalisation Drift in Polarising Topics on TikTok", "published": "2026-03-21T09:13:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-tiktok" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.20062", "title": "The End of Rented Discovery: How AI Search Redistributes Power Between Hotels and Intermediaries", "published": "2026-03-20T15:44:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.19880", "title": "What If Consensus Lies? Selective-Complementary Reinforcement Learning at Test Time", "published": "2026-03-20T11:47:12Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.19710", "title": "AIGQ: An End-to-End Hybrid Generative Architecture for E-commerce Query Recommendation", "published": "2026-03-20T07:27:59Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-taobao", "production-evidence", "search-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B05", "plan_reason": "电商生成、搜索融合与工业排序 — completed" }, { "arxiv_id": "2603.19665", "title": "GenFacet: End-to-End Generative Faceted Search via Multi-Task Preference Alignment in E-Commerce", "published": "2026-03-20T06:01:01Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "search-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.19585", "title": "SaFRO: Satisfaction-Aware Fusion via Dual-Relative Policy Optimization for Short-Video Search", "published": "2026-03-20T02:57:50Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "priority-org-kuaishou", "production-evidence", "search-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B05", "plan_reason": "电商生成、搜索融合与工业排序 — completed" }, { "arxiv_id": "2603.19551", "title": "Learning to Bet for Horizon-Aware Anytime-Valid Testing", "published": "2026-03-20T01:22:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.18953", "title": "Context Bootstrapped Reinforcement Learning", "published": "2026-03-19T14:23:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.18765", "title": "Implicit Grading Bias in Large Language Models: How Writing Style Affects Automated Assessment Across Math, Programming, and Essay Tasks", "published": "2026-03-19T11:20:07Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-alibaba", "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.18620", "title": "Learning to Self-Evolve", "published": "2026-03-19T08:41:24Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.18428", "title": "Adaptive Decoding via Test-Time Policy Learning for Self-Improving Generation", "published": "2026-03-19T02:44:42Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.17947", "title": "Unified Policy Value Decomposition for Rapid Adaptation", "published": "2026-03-18T17:19:56Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.18118", "title": "Insight-V++: Towards Advanced Long-Chain Visual Reasoning with Multimodal Large Language Models", "published": "2026-03-18T15:28:07Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.17168", "title": "HierarchicalKV: A GPU Hash Table with Cache Semantics for Continuous Online Embedding Storage", "published": "2026-03-17T21:59:59Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.13057", "title": "A Multi-Model Approach to English-Bangla Sentiment Classification of Government Mobile Banking App Reviews", "published": "2026-03-17T21:43:51Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.16867", "title": "Efficient Reasoning on the Edge", "published": "2026-03-17T17:59:51Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.16856", "title": "Online Experiential Learning for Language Models", "published": "2026-03-17T17:57:49Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.16755", "title": "A Practical Algorithm for Feature-Rich, Non-Stationary Bandit Problems", "published": "2026-03-17T16:34:23Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-microsoft" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.16500", "title": "From the Inside Out: Progressive Distribution Refinement for Confidence Calibration", "published": "2026-03-17T13:26:29Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.16088", "title": "RecBundle: A Next-Generation Geometric Paradigm for Explainable Recommender Systems", "published": "2026-03-17T03:25:05Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-amazon", "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.15759", "title": "Simulation Distillation: Pretraining World Models in Simulation for Rapid Real-World Adaptation", "published": "2026-03-16T18:00:23Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.15724", "title": "Meta-TTRL: A Metacognitive Framework for Self-Improving Test-Time Reinforcement Learning in Unified Multimodal Models", "published": "2026-03-16T17:28:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.15417", "title": "Amplification Effects in Test-Time Reinforcement Learning: Safety and Reasoning Vulnerabilities", "published": "2026-03-16T15:28:59Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.14333", "title": "Data-Driven Physics Embedded Dynamics with Predictive Control and Reinforcement Learning for Quadrupeds", "published": "2026-03-15T11:52:14Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.14245", "title": "GoldenStart: Q-Guided Priors and Entropy Control for Distilling Flow Policies", "published": "2026-03-15T06:39:09Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.12875", "title": "Test-time RL alignment exposes task familiarity artifacts in LLM benchmarks", "published": "2026-03-13T10:24:19Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.12726", "title": "Anchored Alignment: Preventing Positional Collapse in Multimodal Recommender Systems", "published": "2026-03-13T07:17:09Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.12149", "title": "Linking Perception, Confidence and Accuracy in MLLMs", "published": "2026-03-12T16:47:42Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.12145", "title": "Automatic Generation of High-Performance RL Environments", "published": "2026-03-12T16:45:47Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.11486", "title": "Quantized Inference for OneRec-V2", "published": "2026-03-12T03:13:08Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.11327", "title": "Meta-Reinforcement Learning with Self-Reflection for Agentic Search", "published": "2026-03-11T21:40:26Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.10848", "title": "$V_{0.5}$: Generalist Value Model as a Prior for Sparse RL Rollouts", "published": "2026-03-11T14:57:41Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.10165", "title": "OpenClaw-RL: Train Any Agent Simply by Talking", "published": "2026-03-10T18:59:01Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.09221", "title": "Beyond Test-Time Memory: State-Space Optimal Control for LLM Reasoning", "published": "2026-03-10T05:42:13Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.09195", "title": "$P^2$GNN: Two Prototype Sets to boost GNN Performance", "published": "2026-03-10T05:10:02Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.08987", "title": "MAPLE: Elevating Medical Reasoning from Statistical Consensus to Process-Led Alignment", "published": "2026-03-09T22:22:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "test-time-rl" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.05653", "title": "The DSA's Blind Spot: Algorithmic Audit of Advertising and Minor Profiling on TikTok", "published": "2026-03-05T20:02:48Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-tiktok" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.04816", "title": "Scaling Laws for Cross-Encoder Reranking", "published": "2026-03-05T05:03:07Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.04289", "title": "IPD: Boosting Sequential Policy with Imaginary Planning Distillation in Offline Reinforcement Learning", "published": "2026-03-04T17:05:39Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.04227", "title": "Constraint-Aware Generative Re-ranking for Multi-Objective Optimization in Advertising Feeds", "published": "2026-03-04T16:09:36Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "advertising-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.03988", "title": "SORT: A Systematically Optimized Ranking Transformer for Industrial-scale Recommenders", "published": "2026-03-04T12:32:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B05", "plan_reason": "电商生成、搜索融合与工业排序 — completed" }, { "arxiv_id": "2604.20861", "title": "Deep Interest Mining for Intent-Enriched Semantic IDs in Multimodal Generative Recommendation", "published": "2026-03-03T13:36:22Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.02709", "title": "Sensory-Aware Sequential Recommendation via Review-Distilled Representations", "published": "2026-03-03T08:00:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.02565", "title": "FlashEvaluator: Expanding Search Space with Parallel Sequence-Level Evaluation", "published": "2026-03-03T03:35:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-kuaishou" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.02561", "title": "SOLAR: SVD-Optimized Lifelong Attention for Recommendation", "published": "2026-03-03T03:28:36Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-kuaishou" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.00980", "title": "Beyond the Flat Sequence: Hierarchical and Preference-Aware Generative Recommendations", "published": "2026-03-01T08:15:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.00632", "title": "Stop Treating Collisions Equally: Qualification-Aware Semantic ID Learning for Recommendation at Industrial Scale", "published": "2026-02-28T12:55:49Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "priority-org-kuaishou", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B05", "plan_reason": "电商生成、搜索融合与工业排序 — completed" }, { "arxiv_id": "2603.00502", "title": "Trinity: A Scenario-Aware Recommendation Framework for Large-Scale Cold-Start Users", "published": "2026-02-28T06:58:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-microsoft" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.24125", "title": "Recommendation Algorithms: A Comparative Study in Movie Domain", "published": "2026-02-27T16:01:10Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-netflix" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.23717", "title": "Recommending Search Filters To Improve Conversions At Airbnb", "published": "2026-02-27T06:36:14Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.23530", "title": "Unified Learning-to-Rank for Multi-Channel Retrieval in Large-Scale E-Commerce Search", "published": "2026-02-26T22:26:59Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "search-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.22647", "title": "Vectorizing the Trie: Efficient Constrained Decoding for LLM-based Generative Retrieval on Accelerators", "published": "2026-02-26T06:00:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.22422", "title": "Revisiting Chebyshev Polynomial and Anisotropic RBF Models for Tabular Regression", "published": "2026-02-25T21:34:52Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.21600", "title": "AQR-HNSW: Accelerating Approximate Nearest Neighbor Search via Density-aware Quantization and Multi-stage Re-ranking", "published": "2026-02-25T05:58:16Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "industrial-ranking", "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2603.19249", "title": "Spelling Correction in Healthcare Query-Answer Systems: Methods, Retrieval Impact, and Empirical Evaluation", "published": "2026-02-25T04:58:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.20995", "title": "Generative Pseudo-Labeling for Pre-Ranking with LLMs", "published": "2026-02-24T15:14:49Z", "tracks": [ "recommendation" ], "review_buckets": [ "manual-review" ], "matched_queries": [ "deployment-evidence", "industrial-ranking", "recommendation-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B05", "plan_reason": "电商生成、搜索融合与工业排序 — completed" }, { "arxiv_id": "2603.13253", "title": "A Counterfactual Approach for Addressing Individual User Unfairness in Collaborative Recommender System", "published": "2026-02-24T15:08:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.20877", "title": "E-MMKGR: A Unified Multimodal Knowledge Graph Framework for E-commerce Applications", "published": "2026-02-24T13:19:42Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.20558", "title": "From Logs to Language: Learning Optimal Verbalization for LLM-Based Recommendation at Industry Scale", "published": "2026-02-24T05:15:24Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-netflix" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.06631", "title": "T-REX: Transformer-Based Category Sequence Generation for Grocery Basket Recommendation", "published": "2026-02-23T19:19:46Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-amazon", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.19339", "title": "SplitLight: An Exploratory Toolkit for Recommender Systems Datasets and Splits", "published": "2026-02-22T21:02:32Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.18907", "title": "DeepInterestGR: Mining Deep Multi-Interest Using Multi-Modal LLMs for Generative Recommendation", "published": "2026-02-21T17:03:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.18348", "title": "Explaining AutoClustering: Uncovering Meta-Feature Contribution in AutoML for Clustering", "published": "2026-02-20T17:01:25Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.17976", "title": "In-Context Pure Exploration in Continuous Decision Spaces", "published": "2026-02-20T04:20:47Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.17442", "title": "WarpRec: Unifying Academic Rigor and Industrial Scale for Responsible, Reproducible, and Efficient Recommendation", "published": "2026-02-19T15:09:04Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.17058", "title": "A Long-term Value Prediction Framework In Video Ranking", "published": "2026-02-19T04:01:01Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "industrial-ranking", "priority-org-taobao", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B05", "plan_reason": "电商生成、搜索融合与工业排序 — completed" }, { "arxiv_id": "2602.15508", "title": "Eco-Amazon: Enriching E-commerce Datasets with Product Carbon Footprint for Sustainable Recommendations", "published": "2026-02-17T11:30:11Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.15312", "title": "Extracting Consumer Insight from Text: A Large Language Model Approach to Emotion and Evaluation Measurement", "published": "2026-02-17T02:33:51Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.15005", "title": "Learning User Interests via Reasoning and Distillation for Cross-Domain News Recommendation", "published": "2026-02-16T18:45:40Z", "tracks": [ "post-training", "recommendation" ], "review_buckets": [ "industrial-fulltext-review", "p2-deferred-review" ], "matched_queries": [ "opd", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.13581", "title": "Climber-Pilot: A Non-Myopic Generative Recommendation Model Towards Better Instruction-Following", "published": "2026-02-14T03:46:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.13134", "title": "Awakening Dormant Users: Generative Recommendation with Counterfactual Functional Role Reasoning", "published": "2026-02-13T17:33:48Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-kuaishou", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.12972", "title": "Jointly Optimizing Debiased CTR and Uplift for Coupons Marketing: A Unified Causal Framework", "published": "2026-02-13T14:46:20Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "advertising-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.12968", "title": "RGAlign-Rec: Ranking-Guided Alignment for Latent Query Reasoning in Recommendation Systems", "published": "2026-02-13T14:38:02Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B06", "plan_reason": "2 月召回、广告、长序列与 LLM 排序 — completed" }, { "arxiv_id": "2602.12806", "title": "RAT-Bench: A Comprehensive Benchmark for Text Anonymization", "published": "2026-02-13T10:41:44Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-microsoft" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.12593", "title": "RQ-GMM: Residual Quantized Gaussian Mixture Model for Multimodal Semantic Discretization in CTR Prediction", "published": "2026-02-13T04:11:24Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.12564", "title": "CAPTS: Channel-Aware, Preference-Aligned Trigger Selection for Multi-Channel Item-to-Item Retrieval", "published": "2026-02-13T03:23:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-kuaishou", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.12530", "title": "Reasoning to Rank: An End-to-End Solution for Exploiting Large Language Models for Recommendation", "published": "2026-02-13T02:22:48Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking", "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.12354", "title": "An Industrial-Scale Sequential Recommender for LinkedIn Feed Ranking", "published": "2026-02-12T19:27:15Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "production-evidence", "recommendation-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B06", "plan_reason": "2 月召回、广告、长序列与 LLM 排序 — completed" }, { "arxiv_id": "2602.12041", "title": "Compress, Cross and Scale: Multi-Level Compression Cross Networks for Efficient Scaling in Recommender Systems", "published": "2026-02-12T15:06:46Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.11562", "title": "LASER: An Efficient Target-Aware Segmented Attention Framework for End-to-End Long Sequence Modeling", "published": "2026-02-12T04:33:37Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.11410", "title": "CADET: Context-Conditioned Ads CTR Prediction With a Decoder-Only Transformer", "published": "2026-02-11T22:24:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "advertising-ranking", "deployment-evidence", "industrial-ranking", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B06", "plan_reason": "2 月召回、广告、长序列与 LLM 排序 — completed" }, { "arxiv_id": "2604.20848", "title": "MATRAG: Multi-Agent Transparent Retrieval-Augmented Generation for Explainable Recommendations", "published": "2026-02-11T06:02:31Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence", "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.10490", "title": "ChainRec: An Agentic Recommender Learning to Route Tool Chains for Diverse and Evolving Interests", "published": "2026-02-11T03:50:36Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.10455", "title": "Compute Only Once: UG-Separation for Efficient Large Recommendation Models", "published": "2026-02-11T02:53:59Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-bytedance", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.09901", "title": "QP-OneModel: A Unified Generative LLM for Multi-Task Query Understanding in Xiaohongshu Search", "published": "2026-02-10T15:38:17Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "search-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.09744", "title": "DiffuReason: Bridging Latent Reasoning and Generative Refinement for Sequential Recommendation", "published": "2026-02-10T12:55:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B06", "plan_reason": "2 月召回、广告、长序列与 LLM 排序 — completed" }, { "arxiv_id": "2602.09401", "title": "SARM: LLM-Augmented Semantic Anchor for End-to-End Live-Streaming Ranking", "published": "2026-02-10T04:15:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "industrial-ranking", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B06", "plan_reason": "2 月召回、广告、长序列与 LLM 排序 — completed" }, { "arxiv_id": "2602.09387", "title": "Query-Mixed Interest Extraction and Heterogeneous Interaction: A Scalable CTR Model for Industrial Recommender Systems", "published": "2026-02-10T03:56:14Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.09386", "title": "SMES: Towards Scalable Multi-Task Recommendation via Expert Sparsity", "published": "2026-02-10T03:56:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "priority-org-kuaishou" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.09194", "title": "ML-DCN: Masked Low-Rank Deep Crossing Network Towards Scalable Ads Click-through Rate Prediction at Pinterest", "published": "2026-02-09T20:59:19Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "industrial-ranking", "priority-org-pinterest", "production-evidence" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B06", "plan_reason": "2 月召回、广告、长序列与 LLM 排序 — completed" }, { "arxiv_id": "2602.08837", "title": "AMEM4Rec: Leveraging Cross-User Similarity for Memory Evolution in Agentic LLM Recommenders", "published": "2026-02-09T16:06:55Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.08575", "title": "RankGR: Rank-Enhanced Generative Retrieval with Listwise Direct Preference Optimization in Recommendation", "published": "2026-02-09T12:13:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking", "priority-org-taobao" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.08530", "title": "PIT: A Dynamic Personalized Item Tokenizer for End-to-End Generative Recommendation", "published": "2026-02-09T11:28:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-kuaishou", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.07987", "title": "Learning to Alleviate Familiarity Bias in Video Recommendation", "published": "2026-02-08T14:27:34Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.07774", "title": "GR2: Generative Reasoning Re-ranker", "published": "2026-02-08T02:12:24Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking", "recommendation-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.07521", "title": "Pareto-guided Pipeline for Distilling Featherweight AI Agents in Mobile MOBA Games", "published": "2026-02-07T12:36:38Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2603.12268", "title": "A Holistic Framework for Automated Configuration Recommendation for Cloud Service Monitoring", "published": "2026-02-07T07:54:42Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence", "priority-org-microsoft" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.07208", "title": "Sequences as Nodes for Contrastive Multimodal Graph Recommendation", "published": "2026-02-06T21:35:12Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.07207", "title": "Multimodal Enhancement of Sequential Recommendation", "published": "2026-02-06T21:32:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.05663", "title": "Coarse-to-Fine Long-term Interest Modeling for Generative Recommendation", "published": "2026-02-05T13:48:33Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-taobao" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.04412", "title": "HoRD: Robust Humanoid Control via History-Conditioned Reinforcement Learning and Online Distillation", "published": "2026-02-04T10:41:23Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.04928", "title": "Euphonium: Steering Video Flow Matching via Process Reward Gradient Guided Stochastic Dynamics", "published": "2026-02-04T08:59:57Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.03608", "title": "Controlling Output Rankings in Generative Engines for LLM-based Search", "published": "2026-02-03T14:59:48Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.03324", "title": "SCASRec: A Self-Correcting and Auto-Stopping Model for Generative Route List Recommendation", "published": "2026-02-03T09:51:58Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence", "recommendation-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.02137", "title": "DCoPilot: Generative AI-Empowered Policy Adaptation for Dynamic Data Center Operations", "published": "2026-02-02T14:18:52Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.01058", "title": "Good SFT Optimizes for SFT, Better SFT Prepares for Reinforcement Learning", "published": "2026-02-01T06:53:45Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.01023", "title": "Unifying Ranking and Generation in Query Auto-Completion via Retrieval-Augmented Generation and Multi-Objective Alignment", "published": "2026-02-01T05:15:07Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "industrial-ranking" ], "full_text_review_required": true, "plan_status": "implemented-in-current-pr", "implementation_batch": "B06", "plan_reason": "2 月召回、广告、长序列与 LLM 排序 — completed" }, { "arxiv_id": "2602.00899", "title": "Domain-Adaptive and Scalable Dense Retrieval for Content-Based Recommendation", "published": "2026-01-31T20:58:23Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.02582", "title": "Uncertainty and Fairness Awareness in LLM-Based Recommendation Systems", "published": "2026-01-31T17:18:13Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google", "priority-org-google-deepmind" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.00758", "title": "Temporal Leakage in Search-Engine Date-Filtered Web Retrieval: A Retrospective Forecasting Case Study", "published": "2026-01-31T14:47:01Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-google" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2602.00727", "title": "SWGCN: Synergy Weighted Graph Convolutional Network for Multi-Behavior Recommendation", "published": "2026-01-31T13:42:49Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-taobao" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.00488", "title": "OD-Gear: Online Decomposition and Group Sampling for Expert-Guided Adversarial Routing in Scalable Capacitated Vehicle Routing", "published": "2026-01-31T03:16:54Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.22820", "title": "User-Adaptive Meta-Learning for Cold-Start Medication Recommendation with Uncertainty Filtering", "published": "2026-01-30T10:45:47Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.21452", "title": "SAGE: Sequence-level Adaptive Gradient Evolution for Generative Recommendation", "published": "2026-01-29T09:30:13Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.21285", "title": "Zenith: Scaling up Ranking Models for Billion-scale Livestreaming Recommendation", "published": "2026-01-29T05:36:49Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "priority-org-tiktok", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.20316", "title": "Less is More: Benchmarking LLM Based Recommendation Agents", "published": "2026-01-28T07:08:51Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.20215", "title": "Towards End-to-End Alignment of User Satisfaction via Questionnaire in Video Recommendation", "published": "2026-01-28T03:32:21Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence", "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.20199", "title": "MERGE: Next-Generation Item Indexing Paradigm for Large-Scale Streaming Recommendation", "published": "2026-01-28T02:56:30Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.19965", "title": "Modeling Cascaded Delay Feedback for Online Net Conversion Rate Prediction: Benchmark, Insights and Solutions", "published": "2026-01-27T13:24:32Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking", "priority-org-taobao" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.19121", "title": "LLMs as Orchestrators: Constraint-Compliant Multi-Agent Optimization for Recommendation Systems", "published": "2026-01-27T02:46:13Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.18664", "title": "S$^2$GR: Stepwise Semantic-Guided Reasoning in Latent Space for Generative Recommendation", "published": "2026-01-26T16:40:37Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2604.09549", "title": "Beyond Offline A/B Testing: Context-Aware Agent Simulation for Recommender System Evaluation", "published": "2026-01-26T05:01:00Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.17836", "title": "Unleashing the Potential of Sparse Attention on Long-term Behaviors for CTR Prediction", "published": "2026-01-25T13:39:26Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.17472", "title": "Adversarial Alignment and Disentanglement for Cross-Domain CTR Prediction with Domain-Encompassing Features", "published": "2026-01-24T14:20:16Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.17218", "title": "Evaluation on Entity Matching in Recommender Systems", "published": "2026-01-23T23:05:46Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.16882", "title": "Explaining Group Recommendations via Counterfactuals", "published": "2026-01-23T16:42:05Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.16815", "title": "PI2I: A Personalized Item-Based Collaborative Filtering Retrieval Framework", "published": "2026-01-23T15:10:39Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence", "priority-org-taobao" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2602.02516", "title": "Measuring Individual User Fairness with User Similarity and Effectiveness Disparity", "published": "2026-01-23T14:18:29Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recommendation-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.14333", "title": "Hierarchical Contextual Uplift Bandits for Catalog Personalization", "published": "2026-01-20T11:35:36Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "deployment-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.12263", "title": "Multimodal Generative Engine Optimization: Rank Manipulation for Vision-Language Model Rankers", "published": "2026-01-18T04:58:28Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "recommendation-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.10471", "title": "DeFlow: Decoupling Manifold Modeling and Value Maximization for Offline Policy Extraction", "published": "2026-01-15T14:56:57Z", "tracks": [ "post-training" ], "review_buckets": [ "query-collision" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.10027", "title": "STCRank: Spatio-temporal Collaborative Ranking for Interactive Recommender System at Kuaishou E-shop", "published": "2026-01-15T03:18:40Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "deployment-evidence", "priority-org-kuaishou" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.08421", "title": "Coverage Improvement and Fast Convergence of On-policy Preference Learning", "published": "2026-01-13T10:46:06Z", "tracks": [ "post-training" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "opd" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.08360", "title": "Scalable Sequential Recommendation under Latency and Memory Constraints", "published": "2026-01-13T09:16:49Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.06873", "title": "Applying Embedding-Based Retrieval to Airbnb Search", "published": "2026-01-11T11:41:55Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "search-ranking" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.06458", "title": "PixRec: Leveraging Visual Context for Next-Item Prediction in Sequential Recommendation", "published": "2026-01-10T06:52:58Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2604.06172", "title": "EviSnap: Faithful Evidence-Cited Explanations for Cold-Start Cross-Domain Recommendation", "published": "2026-01-09T18:21:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-amazon" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.04674", "title": "PROMISE: Process Reward Models Unlock Test-Time Scaling Laws in Generative Recommendations", "published": "2026-01-08T07:38:46Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.04554", "title": "Exploring Recommender System Evaluation: A Multi-Modal User Agent Framework for A/B Testing", "published": "2026-01-08T03:33:43Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "production-evidence" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.02955", "title": "Rethinking Multi-objective Ranking Ensemble in Recommender System: From Score Fusion to Rank Consistency", "published": "2026-01-06T11:59:02Z", "tracks": [ "recommendation" ], "review_buckets": [ "industrial-fulltext-review" ], "matched_queries": [ "industrial-ranking", "priority-org-kuaishou" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.02764", "title": "Netflix Artwork Personalization via LLM Post-training", "published": "2026-01-06T06:56:53Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "priority-org-netflix" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" }, { "arxiv_id": "2601.02002", "title": "Exploring Approaches for Detecting Memorization of Recommender System Data in Large Language Models", "published": "2026-01-05T11:03:56Z", "tracks": [ "recommendation" ], "review_buckets": [ "priority-fulltext-review" ], "matched_queries": [ "priority-org-meta" ], "full_text_review_required": true, "plan_status": "fulltext-review-backlog", "implementation_batch": null, "plan_reason": "not in fixed P0/P1 batches; retain until full-text rejection or promotion" }, { "arxiv_id": "2601.01712", "title": "RelayGR: Scaling Long-Sequence Generative Recommendation via Cross-Stage Relay-Race Inference", "published": "2026-01-05T01:34:06Z", "tracks": [ "recommendation" ], "review_buckets": [ "p2-deferred-review" ], "matched_queries": [ "industrial-ranking" ], "full_text_review_required": false, "plan_status": "p2-or-query-collision", "implementation_batch": null, "plan_reason": "below current P0/P1 threshold; retained for audit, not silently discarded" } ] }