arXiv cs.AI· Yipeng Hao, Yuxuan Cai, Wenkai Ji, Jie Zhou, Liyang Yu, Qin Chen, Zhiguo Yan, Shulei Zheng, Liang He·· 2 小时前AI 评分39
DiscernMem：在多智能体系统中精炼记忆并学习何时使用记忆
Remembering Is Not Enough: Refining Memories and Learning When to Use Them in Multi-Agent Systems
AI 导读
研究者提出无需训练的记忆框架 DiscernMem，同时改进 LLM 多智能体系统的记忆质量与使用时机。其 DMR 组件对比有无记忆的团队执行来诊断规则缺陷，BLM 组件依据任务上下文学习各角色应何时应用或避免某条记忆。
正文 · 原文
Abstract:Memory helps LLM-based multi-agent systems reuse experience, but stored advice can be flawed or applied in the wrong situation. Team-level success or failure alone cannot reliably show whether a memory needs correction or should not have been used. We propose DiscernMem, a training-free framework that improves both memory quality and memory use. To improve quality, Defect-Aware Memory Refinement (DMR) compares team executions with a single memory and with no memory, diagnoses rule defects, and tests repairs on current and earlier tasks to guard against losing previously useful behavior. To guide appropriate use, Boundary-Learning Meta-Memory (BLM) records each memory's past uses and effects, learning when its owner role should apply or avoid it based on task context and incoming messages. Together, these components help teams correct flawed experience while retaining useful rules that were applied incorrectly. Experiments with Qwen3-4B-Instruct-2507 on four benchmarks show that DiscernMem achieves the highest average score on AutoGen, CAMEL, and MacNet, exceeding the strongest reported baselines by 0.51-4.32 points. It leads on code generation and symbolic planning across all three systems, including a PDDL score of 37.71 versus 28.48 on CAMEL.
| Subjects: | Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2610.36652 [cs.AI] |
| (or arXiv:2610.36652v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.36652 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yipeng Hao [view email]
[v1]
Tue, 27 Oct 2026 05:12:46 UTC (315 KB)
来源：arXiv cs.AI · arxiv.org