用物理场理论构建连续记忆，让AI长期保持上下文理解力。
Field-Theoretic Memory for AI Agents: Continuous Dynamics for Context Preservation
- 将记忆视为在语义空间中扩散的连续场，通过微分方程动态演化。
- 在500+轮对话中，多轮推理F1提升116%，时间推理准确率提高43.8%。
- 适合需要长期记忆与多智能体协作的复杂任务场景。
我们提出一种面向AI代理的记忆系统，将存储信息视为由偏微分方程控制的连续场，而非数据库中的离散条目。该方法借鉴经典场论：记忆在语义空间中扩散，依重要性热力学衰减，并在多智能体场景中通过场耦合相互作用。我们在两个主流长上下文基准上评估：LoCoMo（ACL 2024）包含35个会话、共300轮对话；LongMemEval（ICLR 2025）测试跨会话推理超过500轮。在LongMemEval上，该方法实现显著提升：多会话推理F1提升116%（p<0.01, d=3.06），时间推理准确率提升43.8%（p<0.001, d=9.21），知识更新检索召回率提升27.8%（p<0.001, d=5.00）。多智能体实验显示，通过场耦合实现接近完美的集体智能（>99.8%）。代码已开源：github.com/rotalabs/rotalabs-fieldmem。
原文摘要 · Abstract (English)
We present a memory system for AI agents that treats stored information as continuous fields governed by partial differential equations rather than discrete entries in a database. The approach draws from classical field theory: memories diffuse through semantic space, decay thermodynamically based on importance, and interact through field coupling in multi-agent scenarios. We evaluate the system on two established long-context benchmarks: LoCoMo (ACL 2024) with 300-turn conversations across 35 sessions, and LongMemEval (ICLR 2025) testing multi-session reasoning over 500+ turns. On LongMemEval, the field-theoretic approach achieves significant improvements: +116% F1 on multi-session reasoning (p<0.01, d= 3.06), +43.8% on temporal reasoning (p<0.001, d= 9.21), and +27.8% retrieval recall on knowledge updates (p<0.001, d= 5.00). Multi-agent experiments show near-perfect collective intelligence (>99.8%) through field coupling. Code is available at github.com/rotalabs/rotalabs-fieldmem.
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。