混合记忆架构通过互补整理与巩固实现长期智能体记忆，LoCoMo 准确率达 90.2%

Long-Term Agent Memory through Complementary Curation and Consolidation

AI 导读

一项研究提出由情景记忆与结构化关系记忆组成的混合记忆架构，并配合巩固机制保持效率。在 LoCoMo 和 LongMemEval-S 两个基准上，两类记忆持续互补，组合后优于现有 SOTA 记忆框架，分别达到 90.2% 和 85.5% 的准确率。

正文 · 原文

View PDF HTML (experimental)

Abstract:Long-horizon dialogue agents require effective memory systems to recall facts, events, and relationships from interaction histories. In this work, we propose a hybrid-memory architecture comprising episodic memories and structured relational memories and a consolidation mechanism which preserves efficiency. We conduct an empirical study across two benchmarks: LoCoMo and LongMemEval-S. Our experiments demonstrate that the two memory types are consistently complementary, combining both outperform state-of-the-art memory frameworks and achieves 90.2% accuracy on LoCoMo and 85.5% on LongMemEval-S.

Subjects:	Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
Cite as:	arXiv:2610.39790 [cs.AI]
	(or arXiv:2610.39790v1 [cs.AI] for this version)
	https://doi.org/10.48550/arXiv.2610.39790 arXiv-issued DOI via DataCite

Submission history

From: Abhishek Gupta [view email]
[v1] Thu, 29 Oct 2026 10:40:18 UTC (603 KB)

来源：arXiv cs.AI · arxiv.org