在大语言模型自进化中通过自适应记忆-参数协调学习该记住什么和该内化什么
Learning What to Remember and What to Internalize in LLM Self-Evolution via Adaptive Memory-Parameter Coordination
- University of Science and Technology of China（中国科学技术大学）
- City University of Hong Kong（香港城市大学）
- Hefei Normal University（合肥师范学院）
- Xi’an Jiaotong University（西安交通大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
该研究提出 COVE 框架，通过任务感知路由等方式协调大语言模型自进化的记忆与参数通道，解决单通道进化的灵活性与性能权衡问题，在多任务上实现更稳健高效的改进。
AI中文摘要：
大语言模型智能体越来越多地在动态环境中运行，其中工具接口、API和用户需求在部署后会发生变化。现有的自进化方法主要遵循两种范式：基于工具 harness 的方法，将反馈外部化为可编辑的记忆或技能以实现快速适应；基于参数的方法，将经验内化到模型参数中以实现更深层次的能力提升。然而，单独使用任一机制都会在灵活性和性能之间产生权衡。本文研究智能体如何协调这两个通道以实现稳健的自进化，提出 COVE 这一统一的智能体自进化框架，通过任务感知路由、阶段感知调度和知识优化结合基于工具 harness 和基于参数的学习。通过该设计，COVE 将自进化视为任务与知识类型匹配适当学习机制的协调过程，而非经验的无差别积累。在多个任务类别上的实验表明，COVE 优于单通道进化策略，在变化环境下展现出更稳健、高效的改进。
英文摘要：
Large language model agents increasingly operate in dynamic environments where tool interfaces, APIs, and user requirements change after deployment. Existing self-evolution methods mainly follow two paradigms: harness-based approaches, which externalize feedback into editable memories or skills for rapid adaptation, and parameter-based approaches, which internalize experience into model parameters for deeper capability improvement. However, using either mechanism alone creates a trade-off between flexibility and performance. This paper asks how an agent can coordinate both channels to achieve robust self-evolution. We present COVE, a unified agent self-evolution framework that combines harness-based and parameter-based learning through task-aware routing, stage-aware scheduling, and knowledge optimization. Through this design, COVE treats self-evolution not as indiscriminate accumulation of experience, but as a coordinated process that matches tasks and knowledge types to appropriate learning mechanisms. Experiments across multiple task categories show that COVE outperforms single-channel evolution strategies, demonstrating more robust and efficient improvement under changing environments.
