arXiv cs.AI· Jian Zhong, Zeyu Liu, Pingchuan Cao, Rongduo Han, Shunye Tang, Guohuan Xie, Mingyuan Qin, Ran Ji, Xiao Liang, Wei Wang, Haining Zhang·· 13 小时前AI 评分38
PreCoMem：面向自进化长期对话智能体的预测式认知记忆框架
PreCoMem: Predictive Cognitive Memory for Self-Evolving Long-Term Dialogue Agents
AI 导读
PreCoMem 是一个受预测编码启发的记忆整合框架，将记忆表示为对用户状态的置信度加权信念，核心信号 Effective Surprisal 结合语义偏差与矛盾证据，并在检索分散时降低其权重。
正文 · 原文
Authors:Jian Zhong, Zeyu Liu, Pingchuan Cao, Rongduo Han, Shunye Tang, Guohuan Xie, Mingyuan Qin, Ran Ji, Xiao Liang, Wei Wang, Haining Zhang
Abstract:Long-horizon dialogue agents need to keep user memories stable enough for personalization while adapting when the user genuinely changes. Existing memory systems often append every observation or overwrite old entries by heuristic rules, which makes transient noise hard to distinguish from real belief drift. We introduce PreCoMem, a memory consolidation framework inspired by predictive coding that represents memory as confidence-weighted beliefs about the user's state. Its core signal, Effective Surprisal, combines semantic deviation and contradiction evidence, then downweights this signal when retrieval is diffuse and unreliable. According to effective surprisal, a two-threshold gate maps each turn to one of three deterministic updates: MAINTAIN reinforces supported beliefs, PROFILE stores ambiguous cues as low-confidence hypotheses, and CORRECT softly decays reliably contradicted facts. The PROFILE stage acts as an explicit waiting room, preventing premature commitment to noisy evidence. Experiments on LoCoMo, LongMemEval-S, and PersonaMem-v2 show consistent state-of-the-art accuracy; our new DiMoBench confirms that PreCoMem adapts to non-stationary belief drift without over-reacting to noise. The code is available at this https URL.
| Comments: | Accepted at NeurIPS 2026 |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.29467 [cs.AI] |
| (or arXiv:2610.29467v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.29467 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jian Zhong [view email]
[v1]
Wed, 21 Oct 2026 09:21:33 UTC (3,902 KB)
来源：arXiv cs.AI · arxiv.org