arXiv cs.CL· Heng Cao, Fan Zhang, Jian Yao, Yujie Zheng, Changlin Zhao, Lu Hao, Yuxuan Wei, Wangze Ni, Huaiyu Fu, Yuqian Sun, Xuyan Mo·· 9 小时前AI 评分34
TAP-PER:用紧凑用户表示实现可扩展的 LLM 个性化
Beyond Retrieval: Learning Compact User Representations for Scalable LLM Personalization
AI 导读
研究者提出 TAP-PER(Temporal Attentive Prefix for PERsonalization),将用户偏好编码为可学习的用户状态前缀嵌入,无需把用户历史序列化进提示词,也无需为每个用户维护重型适配器。
正文 · 原文
Authors:Heng Cao, Fan Zhang, Jian Yao, Yujie Zheng, Changlin Zhao, Lu Hao, Yuxuan Wei, Wangze Ni, Huaiyu Fu, Yuqian Sun, Xuyan Mo
Abstract:Personalizing large language models requires adapting model behavior to individual users while preserving robustness and deployment-scale efficiency. Existing approaches typically personalize LLMs either at the input level, by retrieving user histories or constructing profile prompts, or at the parameter level, by maintaining user-specific parameter-efficient modules. The former makes personalization sensitive to retrieval quality and prompt design, whereas the latter incurs storage and maintenance costs that grow with the user population. To address these limitations, we propose TAP-PER (Temporal Attentive Prefix for PERsonalization), a prefix-based framework that encodes user preferences as learnable representations, avoiding the serialization of user histories into prompts and replacing heavy per-user adapters with lightweight user-state prefix embeddings. Inspired by personalized recommendation systems, TAP-PER decomposes user modeling into user-state and query-conditioned components, and incorporates temporal signals to capture the evolving nature of user interests. Experiments on six LaMP tasks show that TAP-PER consistently outperforms prompt-based and model-based baselines across classification, rating, and generation settings. Moreover, TAP-PER uses 130x fewer per-user parameters than OPPU and roughly half the total parameter footprint of PER-PCS at the 1,000-user scale, demonstrating scalable personalization without prompt-serialized histories or heavy per-user adapters.
| Comments: | EMNLP 2026 Findings |
| Subjects: | Information Retrieval (cs.IR); Computation and Language (cs.CL) |
| Cite as: | arXiv:2606.04547 [cs.IR] |
| (or arXiv:2606.04547v4 [cs.IR] for this version) | |
| https://doi.org/10.48550/arXiv.2606.04547 arXiv-issued DOI via DataCite |
Submission history
From: Heng Cao [view email]
[v1]
Wed, 3 Jun 2026 07:32:18 UTC (433 KB)
[v2]
Sat, 13 Jun 2026 04:11:45 UTC (432 KB)
[v3]
Wed, 2 Sep 2026 03:18:30 UTC (451 KB)
[v4]
Tue, 17 Nov 2026 03:51:50 UTC (442 KB)
来源:arXiv cs.CL · arxiv.org