并非一切尽失：修复个性化编码器的有损用户偏好状态
Not All Is Lost: Repairing Lossy User Preference States of Personalization Encoders
浏览论文内容
中文总结 AI 辅助
提出REPAIR方法，通过比较缓存表示与偏好状态来修复个性化编码器的有损状态，在冻结编码器和任务头的情况下提升推荐和生成性能。
中文摘要 AI 辅助
个性化编码器将不断演化的交互历史压缩为偏好状态，用于对物品进行排序或对文本生成进行条件设定。仅基于该状态运行的任务头可能会遗漏冻结编码器中各时间步缓存表示里仍保留的有用证据。我们研究了这种可恢复性差距，并提出REPAIR方法，该方法在紧凑的学习坐标空间中将缓存表示与当前偏好状态进行比较。它解析来自长期历史、近期交互和局部突发模式中的纠正性证据，然后选择哪些时间步的哪些模式有贡献，并在任务头之前将其聚合修正添加到状态中。编码器宿主修复复用现有前向计算中的表示，无需重新编码历史。在MovieLens、PENS、MIND和Amazon Reviews 2023数据集上，仅训练REPAIR即可在保持编码器和任务头均冻结的情况下，为全部十二个代表性推荐宿主提升MRR和nDCG@10。对相同宿主仅进行头微调所获增益较小。例如，MovieLens上的Mamba4Rec获得了3.96个MRR点的提升，而仅头微调仅提升0.19。排序和时间诊断支持一种紧凑且依赖宿主的纠正结构。在个性化生成中，IMPerSumm将两个已报告的加权PerSEval变体（用于评估对用户偏好的响应性）提升了最高25.23%。这些结果支持压缩后状态修正，并将偏好证据的可用性与其下游使用区分开来。
英文摘要
Personalization encoders compress evolving interaction histories into preference states used to rank items or condition text generation. A task head operating only on this state can miss useful evidence that remains in the frozen encoder's cached representations for individual timesteps. We study this recoverability gap and propose REPAIR, which compares cached representations with the current preference state in a compact learned coordinate space. It resolves corrective evidence over extended history, recent interactions, and localized bursts. It then selects which patterns at which timesteps contribute and adds their aggregate correction to the state before the task head. Encoder-host repair reuses representations from the existing forward computation without re-encoding the history. Across MovieLens, PENS, MIND, and Amazon Reviews 2023, training only REPAIR improves MRR and nDCG@10 for all twelve representative recommendation hosts while both encoder and task head remain frozen. Head-only finetuning of the same hosts yields smaller gains. For example, Mamba4Rec on MovieLens gains 3.96 MRR points, compared with 0.19 from head-only finetuning. Rank and temporal diagnostics support a compact, host-dependent corrective structure. In personalized generation, IMPerSumm improves the two reported weighted PerSEval variants, which assess responsiveness to user preference, by up to 25.23%. These results support post-compression state correction and distinguish the availability of preference evidence from its downstream use.
