构建可追溯的混合记忆系统，提升长时序智能体推理能力
MemWeaver: Weaving Hybrid Memories for Traceable Long-Horizon Agentic Reasoning
- 用三类互联记忆整合长期经验：关系图谱、模式抽象、原始文本
- 多跳和时间推理准确率显著提升，上下文长度减少95%以上
- 适合需要长期记忆与可追溯性的复杂决策场景
基于大语言模型的智能体在长时序交互中需要支持时间一致性、多跳推理和跨会话证据重用的记忆系统。现有方法主要依赖非结构化检索或粗粒度抽象，常导致时间冲突、推理脆弱和可追溯性差。我们提出MemWeaver，一个统一记忆框架，将长期智能体经验整合为三个互连组件：基于时间锚定的图记忆用于结构化关系推理，经验记忆从重复观测中抽象出交互模式，段落记忆保留原始文本证据。MemWeaver采用双通道检索策略，联合获取结构化知识与支持证据，构建紧凑且信息密集的推理上下文。在LoCoMo基准测试中，相比长上下文基线，MemWeaver显著提升多跳和时间推理准确率，同时输入上下文长度减少超过95%。
原文摘要 · Abstract (English)
Large language model-based agents operating in long-horizon interactions require memory systems that support temporal consistency, multi-hop reasoning, and evidence-grounded reuse across sessions. Existing approaches largely rely on unstructured retrieval or coarse abstractions, which often lead to temporal conflicts, brittle reasoning, and limited traceability. We propose MemWeaver, a unified memory framework that consolidates long-term agent experiences into three interconnected components: a temporally grounded graph memory for structured relational reasoning, an experience memory that abstracts recurring interaction patterns from repeated observations, and a passage memory that preserves original textual evidence. MemWeaver employs a dual-channel retrieval strategy that jointly retrieves structured knowledge and supporting evidence to construct compact yet information-dense contexts for reasoning. Experiments on the LoCoMo benchmark demonstrate that MemWeaver substantially improves multi-hop and temporal reasoning accuracy while reducing input context length by over 95\% compared to long-context baselines.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。