X-Rec 技术报告：基于流匹配的连续空间生成式推荐检索
原标题：X-Rec Technical Report
AI 摘要
该论文提出 X-Rec，一种基于流匹配（flow matching）的生成式检索方法，直接在连续物品嵌入空间建模推荐分布，以规避 U2I 表达力不足和 SID-AR 量化误差、低吞吐的问题。X-Rec 引入锚点条件、黎曼流匹配和延迟交互扩散 Transformer 三项设计。在流式基准上，X-Rec 显著优于 U2I 基线，检索质量与 SID-AR 相当，推理吞吐更高。该方法已部署为 TikTok 某垂直内容的检索源，连续两次上线使垂直互动提升 4.1484%、整体互动提升 0.0111%。
正文节选
X-Rec Technical Report Abstract Recent advances in generative modeling have reshaped recommender systems by formulating recommendation as a next-item generation problem. Existing retrieval approaches primarily follow two paradigms: user-to-item (U2I) methods represent user context using one or a few deterministic embeddings, which limits the ability to capture diverse and multi-mode interests, while semantic-ID-based autoregressive (SID-AR) methods model more expressive distributions but suffer