MAGMA：面向AI智能体的多图驱动型代理记忆架构

摘要

记忆增强生成（MAG）通过为大型语言模型引入外部记忆机制来支持长上下文推理，但现有方法主要依赖对单一记忆存储的语义相似度检索，混淆了时序、因果和实体信息。这种设计限制了查询意图与检索证据间的可解释性及对齐能力，导致推理准确度欠佳。本文提出MAGMA——一种多图代理记忆架构，将每个记忆项映射至正交的语义图、时序图、因果图和实体图。MAGMA将检索过程建模为基于策略的多图遍历，实现查询自适应的选择与结构化上下文构建。通过解耦记忆表征与检索逻辑，MAGMA提供了透明的推理路径和细粒度检索控制。在LoCoMo和LongMemEval基准上的实验表明，MAGMA在长程推理任务中持续优于最先进的代理记忆系统。

English

Memory-Augmented Generation (MAG) extends Large Language Models with external memory to support long-context reasoning, but existing approaches largely rely on semantic similarity over monolithic memory stores, entangling temporal, causal, and entity information. This design limits interpretability and alignment between query intent and retrieved evidence, leading to suboptimal reasoning accuracy. In this paper, we propose MAGMA, a multi-graph agentic memory architecture that represents each memory item across orthogonal semantic, temporal, causal, and entity graphs. MAGMA formulates retrieval as policy-guided traversal over these relational views, enabling query-adaptive selection and structured context construction. By decoupling memory representation from retrieval logic, MAGMA provides transparent reasoning paths and fine-grained control over retrieval. Experiments on LoCoMo and LongMemEval demonstrate that MAGMA consistently outperforms state-of-the-art agentic memory systems in long-horizon reasoning tasks.

MAGMA：面向AI智能体的多图驱动型代理记忆架构

MAGMA: A Multi-Graph based Agentic Memory Architecture for AI Agents

摘要

Support