{
  "paper": {
    "arxiv_id": "2606.29946",
    "title": "POEM: Partial-Order Enhanced Real-Time Sequential Modeling for Recommendation",
    "url": "https://arxiv.org/abs/2606.29946",
    "organization": "Kuaishou"
  },
  "dataset": {
    "name": "MovieLens 100K",
    "users": 220,
    "items": 360
  },
  "setup": {
    "adapter": "poem",
    "same_split_and_candidates": true
  },
  "baseline": {
    "name": "transition + content + popularity",
    "hit_at_10": 0.10909090909090909,
    "ndcg_at_10": 0.05400546778896325,
    "fresh_hit_at_10": 0.031746031746031744,
    "head_share_at_10": 0.23772727272727273
  },
  "method": {
    "name": "poem core mechanism (validation blend=0.3)",
    "hit_at_10": 0.10909090909090909,
    "ndcg_at_10": 0.050090001217196894,
    "fresh_hit_at_10": 0.0,
    "head_share_at_10": 0.4090909090909091
  },
  "relative": {
    "hit_at_10_percent": 0.0,
    "ndcg_at_10_percent": -7.250129907339743,
    "fresh_hit_at_10_percent": -100.0,
    "head_share_at_10_percent": 72.08413001912048
  },
  "stages": {
    "finite_scores": 360,
    "score_std": 2.085498114162523,
    "ranking_signals": 3,
    "partial_order_pairs": 64620
  },
  "paper_results": {
    "main_usage_time_per_user_lift_percent": 0.249,
    "lite_usage_time_per_user_lift_percent": 0.213
  },
  "scope": "执行论文可由公开 MovieLens 特征审计的核心计算；不复刻私有日志、生产大模型、多媒体生成器或线上服务栈。",
  "schema_version": 2,
  "manifest": {
    "adapter_key": "poem",
    "arxiv_id": "2606.29946",
    "title": "POEM: Partial-Order Enhanced Real-Time Sequential Modeling for Recommendation",
    "paper_url": "https://arxiv.org/abs/2606.29946",
    "track": "recommendation",
    "organization": "Kuaishou",
    "published": "2026-06-29",
    "code_url": null,
    "topics": [
      "sequential-recommendation",
      "partial-order",
      "real-time"
    ],
    "local_code_dir": "src/auto_research/reproductions/industrial_2026",
    "fidelity": "core_mechanism",
    "evaluation_tier": "l2_public_dataset",
    "datasets": [
      "MovieLens 100K"
    ],
    "baseline": "chronological sequential ranking",
    "metrics": [
      "hit_at_10",
      "ndcg_at_10",
      "ranking_signals",
      "partial_order_pairs"
    ],
    "default_seeds": [
      42,
      43,
      44
    ],
    "budget": "220 users / 360 items; validation-only blend selection",
    "device_capabilities": [
      "cpu",
      "mps",
      "cuda"
    ],
    "requires_gpu_validation": false,
    "gpu_validation_artifact": null,
    "online_evidence": [
      {
        "product": "Kuaishou main page",
        "metric": "Usage Time per User",
        "lift_percent": 0.249,
        "traffic": "seven-day test; 5% treatment and 5% control",
        "source_url": "https://arxiv.org/html/2606.29946v1",
        "source_location": "Section 4.2.3, Table 5",
        "retrieved_at": "2026-09-02"
      }
    ],
    "selection_exception": null,
    "evolve_operators": []
  },
  "provenance": {
    "created_at": "2026-09-01T16:16:57.743010+00:00",
    "code_commit": "1d956119a5ece6c757bd4acbc1ed0935420a7915",
    "python": "3.12.9",
    "platform": "sanitized public CPU run",
    "dataset_dir": "data",
    "dataset_fingerprint": "e66c261317a7b3179720ef3e3f5f79d7a204b135d1f217e2911b153ad2ce50a0",
    "packages": {
      "auto-research": "0.1.0",
      "numpy": "2.5.2",
      "torch": "2.13.0",
      "transformers": "5.14.1"
    },
    "artifact_path": "docs/reproductions/2606.29946-poem/metrics/public-seeds42-44.json"
  },
  "evaluation_protocol": {
    "tier": "l2_public_dataset",
    "tier_label": "L2 公开数据集训练",
    "seeds": [
      42,
      43,
      44
    ],
    "budget": "220 users / 360 items; validation-only blend selection",
    "formal_comparison": true,
    "claim_policy": "formal multi-seed comparison"
  },
  "seed_results": [
    {
      "paper": {
        "arxiv_id": "2606.29946",
        "title": "POEM: Partial-Order Enhanced Real-Time Sequential Modeling for Recommendation",
        "url": "https://arxiv.org/abs/2606.29946",
        "organization": "Kuaishou"
      },
      "dataset": {
        "name": "MovieLens 100K",
        "users": 220,
        "items": 360
      },
      "setup": {
        "adapter": "poem",
        "same_split_and_candidates": true
      },
      "baseline": {
        "name": "transition + content + popularity",
        "hit_at_10": 0.10909090909090909,
        "ndcg_at_10": 0.05400546778896325,
        "fresh_hit_at_10": 0.031746031746031744,
        "head_share_at_10": 0.23772727272727273
      },
      "method": {
        "name": "poem core mechanism (validation blend=0.3)",
        "hit_at_10": 0.10909090909090909,
        "ndcg_at_10": 0.050090001217196894,
        "fresh_hit_at_10": 0.0,
        "head_share_at_10": 0.4090909090909091
      },
      "relative": {
        "hit_at_10_percent": 0.0,
        "ndcg_at_10_percent": -7.250129907339743,
        "fresh_hit_at_10_percent": -100.0,
        "head_share_at_10_percent": 72.08413001912048
      },
      "stages": {
        "finite_scores": 360,
        "score_std": 2.085498114162523,
        "ranking_signals": 3,
        "partial_order_pairs": 64620
      },
      "paper_results": {
        "main_usage_time_per_user_lift_percent": 0.249,
        "lite_usage_time_per_user_lift_percent": 0.213
      },
      "scope": "执行论文可由公开 MovieLens 特征审计的核心计算；不复刻私有日志、生产大模型、多媒体生成器或线上服务栈。"
    },
    {
      "paper": {
        "arxiv_id": "2606.29946",
        "title": "POEM: Partial-Order Enhanced Real-Time Sequential Modeling for Recommendation",
        "url": "https://arxiv.org/abs/2606.29946",
        "organization": "Kuaishou"
      },
      "dataset": {
        "name": "MovieLens 100K",
        "users": 220,
        "items": 360
      },
      "setup": {
        "adapter": "poem",
        "same_split_and_candidates": true
      },
      "baseline": {
        "name": "transition + content + popularity",
        "hit_at_10": 0.10909090909090909,
        "ndcg_at_10": 0.05400546778896325,
        "fresh_hit_at_10": 0.031746031746031744,
        "head_share_at_10": 0.23772727272727273
      },
      "method": {
        "name": "poem core mechanism (validation blend=0.3)",
        "hit_at_10": 0.10909090909090909,
        "ndcg_at_10": 0.050090001217196894,
        "fresh_hit_at_10": 0.0,
        "head_share_at_10": 0.4090909090909091
      },
      "relative": {
        "hit_at_10_percent": 0.0,
        "ndcg_at_10_percent": -7.250129907339743,
        "fresh_hit_at_10_percent": -100.0,
        "head_share_at_10_percent": 72.08413001912048
      },
      "stages": {
        "finite_scores": 360,
        "score_std": 2.085498114162523,
        "ranking_signals": 3,
        "partial_order_pairs": 64620
      },
      "paper_results": {
        "main_usage_time_per_user_lift_percent": 0.249,
        "lite_usage_time_per_user_lift_percent": 0.213
      },
      "scope": "执行论文可由公开 MovieLens 特征审计的核心计算；不复刻私有日志、生产大模型、多媒体生成器或线上服务栈。"
    },
    {
      "paper": {
        "arxiv_id": "2606.29946",
        "title": "POEM: Partial-Order Enhanced Real-Time Sequential Modeling for Recommendation",
        "url": "https://arxiv.org/abs/2606.29946",
        "organization": "Kuaishou"
      },
      "dataset": {
        "name": "MovieLens 100K",
        "users": 220,
        "items": 360
      },
      "setup": {
        "adapter": "poem",
        "same_split_and_candidates": true
      },
      "baseline": {
        "name": "transition + content + popularity",
        "hit_at_10": 0.10909090909090909,
        "ndcg_at_10": 0.05400546778896325,
        "fresh_hit_at_10": 0.031746031746031744,
        "head_share_at_10": 0.23772727272727273
      },
      "method": {
        "name": "poem core mechanism (validation blend=0.3)",
        "hit_at_10": 0.10909090909090909,
        "ndcg_at_10": 0.050090001217196894,
        "fresh_hit_at_10": 0.0,
        "head_share_at_10": 0.4090909090909091
      },
      "relative": {
        "hit_at_10_percent": 0.0,
        "ndcg_at_10_percent": -7.250129907339743,
        "fresh_hit_at_10_percent": -100.0,
        "head_share_at_10_percent": 72.08413001912048
      },
      "stages": {
        "finite_scores": 360,
        "score_std": 2.085498114162523,
        "ranking_signals": 3,
        "partial_order_pairs": 64620
      },
      "paper_results": {
        "main_usage_time_per_user_lift_percent": 0.249,
        "lite_usage_time_per_user_lift_percent": 0.213
      },
      "scope": "执行论文可由公开 MovieLens 特征审计的核心计算；不复刻私有日志、生产大模型、多媒体生成器或线上服务栈。"
    }
  ],
  "aggregate_metrics": {
    "dataset.users": {
      "mean": 220.0,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "dataset.items": {
      "mean": 360.0,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "baseline.hit_at_10": {
      "mean": 0.10909090909090909,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "baseline.ndcg_at_10": {
      "mean": 0.05400546778896325,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "baseline.fresh_hit_at_10": {
      "mean": 0.031746031746031744,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "baseline.head_share_at_10": {
      "mean": 0.2377272727272727,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "method.hit_at_10": {
      "mean": 0.10909090909090909,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "method.ndcg_at_10": {
      "mean": 0.050090001217196894,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "method.fresh_hit_at_10": {
      "mean": 0.0,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "method.head_share_at_10": {
      "mean": 0.4090909090909091,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "relative.hit_at_10_percent": {
      "mean": 0.0,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "relative.ndcg_at_10_percent": {
      "mean": -7.250129907339743,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "relative.fresh_hit_at_10_percent": {
      "mean": -100.0,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "relative.head_share_at_10_percent": {
      "mean": 72.08413001912048,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "stages.finite_scores": {
      "mean": 360.0,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "stages.score_std": {
      "mean": 2.085498114162523,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "stages.ranking_signals": {
      "mean": 3.0,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "stages.partial_order_pairs": {
      "mean": 64620.0,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "paper_results.main_usage_time_per_user_lift_percent": {
      "mean": 0.249,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    },
    "paper_results.lite_usage_time_per_user_lift_percent": {
      "mean": 0.213,
      "std": 0.0,
      "ci95": 0.0,
      "n": 3
    }
  },
  "manifest_ref": "reproduction:poem"
}