让大模型像人一样用联想记忆，解决长期记忆断连问题
SYNAPSE: Empowering LLM Agents with Episodic-Semantic Memory via Spreading Activation
- 用动态图结构模拟记忆联想，通过激活扩散实现语义关联
- 在LoCoMo基准上显著超越现有方法，多跳推理准确率提升12.7%
- 适合需要长期记忆与复杂推理的智能体开发场景
尽管大型语言模型在通用推理方面表现优异，但传统检索增强方法无法解决长期代理记忆的离散性问题。为此，我们提出Synapse（协同关联处理语义编码），一种统一的记忆架构，突破了静态向量相似性的限制。借鉴认知科学，Synapse将记忆建模为动态图，相关性通过激活扩散而非预计算链接产生。通过引入横向抑制和时间衰减机制，系统可动态突出相关子图并过滤干扰。我们实现了三重混合检索策略，融合几何嵌入与基于激活的图遍历。在LoCoMo基准上的全面评估表明，Synapse在复杂的时间性和多跳推理任务中显著优于当前最优方法，有效解决了‘上下文隧道’问题。代码与数据将在论文录用后公开。
原文摘要 · Abstract (English)
While Large Language Models (LLMs) excel at generalized reasoning, standard retrieval-augmented approaches fail to address the disconnected nature of long-term agentic memory. To bridge this gap, we introduce Synapse (Synergistic Associative Processing Semantic Encoding), a unified memory architecture that transcends static vector similarity. Drawing from cognitive science, Synapse models memory as a dynamic graph where relevance emerges from spreading activation rather than pre-computed links. By integrating lateral inhibition and temporal decay, the system dynamically highlights relevant sub-graphs while filtering interference. We implement a Triple Hybrid Retrieval strategy that fuses geometric embeddings with activation-based graph traversal. Comprehensive evaluations on the LoCoMo benchmark show that Synapse significantly outperforms state-of-the-art methods in complex temporal and multi-hop reasoning tasks, offering a robust solution to the "Contextual Tunneling" problem. Our code and data will be made publicly available upon acceptance.
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。