发表机构
Tsinghua University; Shenzhen International Graduate School（清华大学; 深圳国际研究生院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究针对长期对话智能体的记忆问题，提出基于PEC²F图模式的CogMem认知记忆架构，结合图操作符重构查询相关上下文，在LoCoMo和LongMemEval数据集上的多跳等任务中表现优异。
AI 中文摘要
作为长期对话智能体的大语言模型（LLM）需要能够对扩展交互进行可靠推理的记忆系统。然而，现有的检索增强生成（RAG）框架通常将记忆视为被动存储，难以区分带有来源属性的信念与无来源的事件/事实记录，也难以连接分散在不同会话中的证据。我们提出CogMem，一种基于PEC²F（人物-事件-概念-主张-事实）图模式的认知记忆架构。专门的主张（Claim）节点保留主观陈述的来源和目标，而事实（Fact）和事件（Event）节点则分别表示语义和情景知识。对话轮次被逐步转换为带有来源信息的图记录，合并为更高级别的事实，并在同一来源提供冲突更新时，协调为具有时间范围的主张视图。对于检索，由LLM意图解析驱动的基于规则的控制器组合了四个确定性图操作符——锚定（anchoring）、遍历（traversal）、交集（intersection）和证据接地（evidence grounding）——以重构与查询相关的上下文。在LoCoMo和LongMemEval上的实验显示出强劲性能，尤其在多跳、时间和知识更新任务上表现突出。消融实验和语义崩溃探测验证了认知分离、合并和智能体检索的互补贡献。代码：this https URL。
英文摘要
Large Language Models (LLMs) serving as long-term dialogue agents require memory systems that support reliable reasoning over extended interactions. However, existing Retrieval-Augmented Generation (RAG) frameworks typically treat memory as passive storage, making it difficult to distinguish source-attributed beliefs from unattributed event/fact records and to connect evidence dispersed across sessions. We introduce CogMem, a cognitive memory architecture based on the PEC$^2$F (Person-Event-Concept-Claim-Fact) graph schema. Dedicated Claim nodes preserve the source and target of subjective statements, while Fact and Event nodes represent semantic and episodic knowledge. Dialogue turns are incrementally converted into provenance-aware graph records, consolidated into higher-level facts, and reconciled into temporally scoped Claim views when the same source provides conflicting updates. For retrieval, a rule-based controller driven by LLM intent parsing composes four deterministic graph operators---anchoring, traversal, intersection, and evidence grounding---to reconstruct query-relevant context. Experiments on LoCoMo and LongMemEval show strong performance, especially on multi-hop, temporal, and knowledge-update tasks. Ablations and a semantic-collapse probe support complementary contributions from epistemic separation, consolidation, and agentic retrieval. Code: https://github.com/Silent-Rain02/CogMem.
CommentsAccepted to EMNLP 2026 (main conference). 21 pages
