让AI通过记忆实时学习新技能，不丢旧知识也不需重训
MemRL: Self-Evolving Agents via Runtime Reinforcement Learning on Episodic Memory
- 用运行时记忆做强化学习，不改模型参数
- 在多个任务中表现优于现有方法，持续提升性能
- 适合需要长期自我改进的智能体系统
人类智能的核心是通过过往经验不断掌握新技能。当前AI代理难以实现这种自我进化：微调成本高且易产生灾难性遗忘，而现有基于记忆的方法依赖被动语义匹配，常检索到噪声信息。为此，我们提出MemRL，一种基于情景记忆的非参数化自进化方法。通过将稳定推理与可塑记忆分离，MemRL采用双阶段检索机制，利用环境反馈过滤噪声，识别高价值策略。在HLE、BigCodeBench、ALFWorld和Lifelong Agent Bench上的大量实验表明，MemRL显著优于当前最优基线，证实其有效解决了稳定性-可塑性矛盾，实现无需权重更新的持续运行时优化。代码已开源：https://github.com/MemTensor/MemRL。
原文摘要 · Abstract (English)
The hallmark of human intelligence is the self-evolving ability to master new skills by learning from past experiences. However, current AI agents struggle to emulate this self-evolution: fine-tuning is computationally expensive and prone to catastrophic forgetting, while existing memory-based methods rely on passive semantic matching that often retrieves noise. To address these challenges, we propose MemRL, a non-parametric approach that evolves via reinforcement learning on episodic memory. By decoupling stable reasoning from plastic memory, MemRL employs a Two-Phase Retrieval mechanism to filter noise and identify high-utility strategies through environmental feedback. Extensive experiments on HLE, BigCodeBench, ALFWorld, and Lifelong Agent Bench demonstrate that MemRL significantly outperforms state-of-the-art baselines, confirming that MemRL effectively reconciles the stability-plasticity dilemma, enabling continuous runtime improvement without weight updates. Code is available at https://github.com/MemTensor/MemRL.
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。