MemoWM：世界模型如何改变智能体需要记忆的内容

中文摘要 AI 辅助

长期智能体在积累经验时会面临不断增长的存储需求。世界模型可捕捉可复用的规律，从而减少为每条经验存储的信息。我们提出了基于世界模型的内存分配问题，并引入MemoWM，这是一个利用共享预测来压缩保留信息并重构被省略内容的框架。其任务感知分配规则在重构误差的预期影响与存储成本之间取得平衡，保留具有超出预测先验的下游价值的信息。在5个长期智能体-内存基准测试中，MemoWM的平均答案准确率为42.42%，比最强基线高出2.62个百分点；与存储效率最高的基线MIRIX相比，其平均特定经验存储量减少了53.9%。进一步分析表明，在可比任务质量下，更强的世界模型可减少每条经验的存储量。考虑模型参数后，共享模型容量与 recurring 存储成本之间存在权衡，随着保留的交互增多，使总存储量最小的容量也会增加。我们的代码可在此https URL获取。

英文摘要

Long-term agents face growing storage demands as they accumulate experience. World models capture reusable regularities that can reduce the information stored for each experience. We formulate the problem of memory allocation conditioned on a world model and introduce MemoWM, a framework that uses shared predictions to compress retained information and reconstruct omitted content. Its task-aware allocation rule balances the expected impact of reconstruction errors against storage cost, retaining information with downstream value beyond the predictive prior. Across five long-term agent-memory benchmarks, MemoWM achieves 42.42\% average answer accuracy, exceeding the strongest baseline by 2.62 percentage points, while reducing average experience-specific storage by 53.9\% relative to MIRIX, the most storage-efficient baseline. Further analysis shows that stronger world models reduce per-experience storage at comparable task quality. Accounting for model parameters reveals a trade-off between shared model capacity and recurring storage costs, with the capacity that minimizes total storage increasing as more interactions are retained. Our code is available at https://github.com/Feld-maxiu/MemoWM.

发表机构

• South China University of Technology（华南理工大学）
• University of the Chinese Academy of Sciences（中国科学院大学）
• Beijing University of Posts and Telecommunications（北京邮电大学）
• University of Science and Technology of China（中国科学技术大学）
• Tsinghua University（清华大学）
• Peking University（北京大学）
• Fudan University（复旦大学）
• Shanghai Jiao Tong University（上海交通大学）

机构由 AI 辅助整理，请以论文原文为准。