知识、协同向量与语义码：相同内容怎样进入不同推荐接口？¶
搜广推 · 更新 2026-10-27 · 60 篇主题资料 / 4 篇重点对照
KAR生成外部知识后适配既有推荐器，NoteLLM用共读监督训练可检索压缩状态，Dual-purpose SID让短码兼顾身份与内容重建，FineSID则保留细粒度token和完整交互骨干做阶段迁移。它们不只是“都用LLM或SID”：对齐信号、压缩对象与下游继承内容各不相同。
机制对照 · 取舍分析 · 跨论文问答 · 更新记录 · 本主题逐篇问答
怎么读这张卡片
重点比较下表代表论文；完整主题目录保留在文末。机制与假设的对照不等于统一 benchmark 的性能排名；教学例子不是论文实测。
机制对照¶
比较维度：输入知识来自生成常识、内容还是交互关系 · 传递对象是向量、短码还是完整预测器状态 · 内容重建与个性化目标如何共同约束 · 离线缓存、冷启动与完整复现边界
| 论文与来源 | 机制定位 | 得到了什么，付出什么 |
|---|---|---|
| Towards Open-World Recommendation with Knowledge Augmentation from Large Language Models · 原始来源 章节：背景与主要改动、核心公式、本地复现 |
分别生成用户偏好推理知识和物品事实知识，以hybrid-expert adapter压缩，再与原ID表示一起送入推荐模型。 | 可预生成知识以移出在线自回归成本，但生成描述不是真实偏好标签，缓存也有刷新与事实核验成本；并非LLM直接在线做全部推荐。 |
| NoteLLM: A Retrievable Large Language Model for Note Recommendation · 原始来源 章节：背景与主要改动、核心公式、本地复现 |
compression prompt把内容压到特殊token状态，GCL用共读关系学习检索空间，CSFT从同一状态生成hashtag/category。 | 协同对比与生成监督共同塑造表示，仍需处理共读关系的曝光偏差；内容相似和行为互补并非天然相同，生成能力也不能只凭检索分数推断。 |
| Tokens are All You Need: Dual-purpose Semantic IDs for Achieving LLM-Level I/O Efficiency in Recommendation Systems · 原始来源 章节：背景与主要改动、核心公式、复现边界 |
层次SID承载协同身份，并通过SiDec从各层离散码重建内容embedding，供召回、排序和低I/O重建使用。 | 短码与重建降低内容传输需求，也引入量化碰撞和信息瓶颈；重建MSE与最终推荐必须分测。本地genre替代多模态内容，未复刻生产码本与在线重建。 |
| Progressive Learning from Content to Relevance to Personalization for CTR Prediction · 原始来源 章节：背景与主要改动、三阶段结构、成本判定：不执行复现 |
保留多尺度视觉/字段文本token，先训练完整query-item相关性预测器，再保留共享dense、embedding与归一化状态并切换CTR头。 | 迁移对象比单个embedding丰富，但要求阶段一致的骨干和状态管理；不等于只接SID lookup。本库仅方法及线上证据收录，成本过高未实现完整训练。 |
取舍分析¶
先问传过去的是什么¶
KAR传递的是生成知识经adapter得到的特征，NoteLLM传递协同训练过的压缩向量，Dual-purpose SID传递可重建内容的离散身份，FineSID则连交互预测器的共享参数与归一化状态一起迁移。若只给既有CTR模型加一张随机码表，既没有KAR的知识来源，也没有FineSID的阶段继承，不能凭接口同为token就视作等价。
对照来源：kar · notellm · dual-sid · finesid
内容保真与推荐相关性需要两把尺¶
Dual-purpose SID的内容重建项检验码里能还原多少embedding，NoteLLM的GCL则用共读关系调整哪些内容应相近；二者约束的几何不同。FineSID再用相关性软目标和个性化CTR分阶段训练同一骨干。更低重建误差不自动带来更好的点击排序，LLM说得更像也不等于用户更喜欢。
对照来源：dual-sid · notellm · finesid
跨论文问答¶
先想自己的答案，再展开；每个问题至少联系两篇论文。
Dual-purpose SID和FineSID都用离散token，为什么不能用同一个SID lookup实验同时证明两篇？¶
展开答案与例子
答案： Dual-purpose SID要求短码同时服务协同身份与内容重建；FineSID还要求细粒度、多尺度位置保留，并把相关性阶段的完整共享预测器迁移到CTR。单个lookup只能覆盖接口的一小部分，缺少各自重建目标或阶段继承，就没有验证其定义性机制。
教学例子： 示意：给鞋图分配三个码后能还原大致内容，是一个压缩问题；先学query—鞋图相关性，再继承整套交互网络预测某用户点击，是另一条训练链路。
KAR生成的偏好描述为什么不能替代NoteLLM使用的共读监督？¶
展开答案与例子
答案： KAR提供常识和偏好推理的候选知识，是否有用还要由推荐目标学习；NoteLLM的共读pair来自行为关系，直接约束检索空间。前者可能补足日志没有的语义，也可能生成错误解释；后者有真实交互依据，却受曝光和流行度影响。两种信号可互补，不能互当真值。
教学例子： 示意：LLM说露营者可能喜欢登山鞋，只是一条知识假设；用户实际常把某双鞋与防水袜一起阅读，则是平台行为关系，二者证据性质不同。
SID重建更准确、文本生成更流畅，为什么推荐NDCG仍可能下降？¶
展开答案与例子
答案： 重建保留内容，CSFT保留生成接口，但排序还需要用户—物品的协同区分。Dual-purpose SID码本可能更偏重内容而牺牲偏好，NoteLLM要靠GCL补充行为几何，FineSID则继续训练个性化CTR。应分别看重建、碰撞、相关性和隔离测试排序，不能把某一辅助目标当最终效果。
教学例子： 示意：两件外观近似的外套被重建得很准，但一个适合骑行、一个适合通勤；若码没有保留用户群差异，视觉保真并不能保证点击次序正确。
更新记录¶
新论文带来新路线、改变取舍，或补充现有路线时，都在这里说明。
2026-10-27 · 首版：比较kar、notellm、dual-sid、finesid，围绕输入知识来自生成常识、内容还是交互关系与传递对象是向量、短码还是完整预测器状态建立分析和跨论文问答。
覆盖 60 篇主题资料；目录见下。
本主题全部 60 篇资料
目录归类覆盖本主题全部资料；重点对照范围以上表为准。
- Beyond Static Collision Handling: Adaptive Semantic ID Learning for Multimodal Recommendation at Industrial Scale
- AgentWebRec: Compact Evidence Fusion over the Agent Web for Personalized Recommendation
- From Head to Tail: Asymmetric Knowledge Transfer in Long-tail Recommendation with Generative Semantic IDs
- Atomic Intent Reasoning: Bringing LLM Semantics to Industrial Cross-Domain Recommendations
- Breaking the Length Barrier: LLM-Enhanced CTR Prediction in Long Textual User Behaviors
- CAMIE: Co-Engagement-Aware Multimodal Item Embeddings for Snap Dynamic Product Ads Retrieval
- ConnectionMind: A General Social-Personalized Recommendation System with LLM Reasoning
- Zero-shot Cross-domain Knowledge Distillation: A Case study on YouTube Music
- Douyin Multimodal Embedding Model Technical Report
- DOS: Dual-Flow Orthogonal Semantic IDs for Recommendation in Meituan
- Tokens are All You Need: Dual-purpose Semantic IDs for Achieving LLM-Level I/O Efficiency in Recommendation Systems
- From a Static Multi-Level Small Semantic Codebook to a Dynamic Single-Level Large Semantic Codebook for Generative Recommendation
- FilterLLM: Text-To-Distribution LLM for Billion-Scale Cold-Start Recommendation
- Progressive Learning from Content to Relevance to Personalization for CTR Prediction
- FLUID: From Ephemeral IDs to Multimodal Semantic Codes for Industrial-Scale Livestreaming Recommendation
- Structuring and Tokenizing Distributed User Interest Context for Generative Recommendation
- GALA: Generative Aligned Learning for Adaptive Multimodal Representation in the Taobao Shangou Recommender System
- GateSID: Adaptive Gating for Semantic-Collaborative Alignment in Cold-Start Recommendation
- Neither Black nor White: Balancing Semantic and Collaborative Signals with Graph-Informed Semantic IDs (GrIS)
- IDProxy: Cold-Start CTR Prediction for Ads and Recommendation at Xiaohongshu with Multimodal LLMs
- Towards Open-World Recommendation with Knowledge Augmentation from Large Language Models
- Knowledge–Geometry Decoupling: Refreshable Pretrained Transfer for Streaming Recommendation
- KuaFu: Compressing Long User Behavior into Understanding at Billion Scale
- L2Rec: Towards Dual-View Understanding of LLMs for Personalized Recommendation
- Knowledge Adaptation from Large Language Model to Recommendation for Practical Industrial Application
- A Large Language Model Enhanced Sequential Recommender for Joint Video and Comment Recommendation
- Unlocking Scaling Law in Industrial Recommendation Systems with a Three-step Paradigm based Large User Model
- LWGR: Lagrangian-Constrained Personalized World Knowledge for Generative Recommendation
- M6-Rec: Generative Pretrained Language Models are Open-Ended Recommender Systems
- Modular Representation Compression: Adapting LLMs for Efficient and Effective Recommendations
- MIM: Multi-modal Content Interest Modeling Paradigm for User Behavior Modeling
- A General Framework for Multimodal LLM-Based Multimedia Understanding in Large-Scale Recommendation Systems
- Mosaic: A Fleet of User Embedding Specialists for Recommendation at Meta
- Balancing Efficiency and Effectiveness: An LLM-Infused Approach for Optimized CTR Prediction
- MuChator: Enabling Active Music Discovery via Conversational Music LLMs in Douyin Music
- Explainable Recommendations at Scale: LLM Rationales for YouTube Music Artist Discovery
- Multimedia Asset Personalization via Multimodal Embeddings at Netflix
- NoteLLM: A Retrievable Large Language Model for Note Recommendation
- Building a User Foundation Model for the Open Web
- Pin-SCALE: Semantic Cascading and Alignment Learning for Engagement-Aware IDs in Cold-Start Recommendations
- PinCLIP: Large-scale Foundational Multimodal Representation at Pinterest
- Fine-Tuned LLM as a Complementary Predictor Improving Ads System
- PIT: A Dynamic Personalized Item Tokenizer for End-to-End Generative Recommendation
- PRECISE: Pre-training Sequential Recommenders with Collaborative and Semantic Information
- Proximity Features: Privacy-Compliant Cold-Start Personalization at Airbnb
- Stop Treating Collisions Equally: Qualification-Aware Semantic ID Learning for Recommendation at Industrial Scale
- RECAP: Feedback-Driven Streaming Semantic User Profiles for Short-Video Recommendation
- RecGPT-Mobile: On-Device Large Language Models for User Intent Understanding in Taobao Feed Recommendation
- RecGPT-V2 Technical Report
- RecoReward: Recommender-Guided Multimodal Description Generation for Recommendation
- RQ-GMM: Residual Quantized Gaussian Mixture Model for Multimodal Semantic Discretization in CTR Prediction
- SARM: LLM-Augmented Semantic Anchor for End-to-End Live-Streaming Ranking
- SaviorRec: Semantic-Behavior Alignment for Cold-Start Recommendation
- Bursting Filter Bubble: Enhancing Serendipity Recommendations with Aligned Large Language Models
- SID-Coord: Coordinating Semantic IDs for ID-based Ranking in Short-Video Search
- Don't Rewrite the SID Map: Topology Preservation for Generative Retrieval in Dynamic E-Commerce Search
- SIRF: A Spec-Internalized Risk Foundation Model for Industrial Content Risk Control
- SSRLive: Live Streaming Recommendation with Dynamic Semantic ID
- TagLLM: A Fine-Grained Tag Generation Approach for Note Recommendation
- TokenMinds: Pretrained User Tokens and Embeddings for User Understanding in Large Recommender Systems