MemEvolve：智能體記憶系統的元演化
MemEvolve: Meta-Evolution of Agent Memory Systems
December 21, 2025
作者: Guibin Zhang, Haotian Ren, Chong Zhan, Zhenhong Zhou, Junhao Wang, He Zhu, Wangchunshu Zhou, Shuicheng Yan
cs.AI
摘要
自我演化記憶系統正以前所未有的方式重塑基於大型語言模型（LLM）的智能體演化範式。現有研究主要依賴人工設計的記憶架構來儲存軌跡、提煉經驗並合成可重用工具，使智能體能在環境互動中實時演化。然而，這種範式本質上受制於記憶系統自身的靜態性：雖然記憶促進了智能體層面的演化，但其底層記憶架構無法針對多樣化任務情境進行元適應。為解決這一侷限，我們提出MemEvolve——一個元演化框架，能同步演化智能體的經驗知識與記憶架構，使智能體系統不僅能積累經驗，更能持續優化其從經驗中學習的方式。為將MemEvolve紮根於既有研究並促進未來自我演化系統的開放性，我們構建了EvolveLab：一個統一的自我演化記憶代碼庫，將十二種代表性記憶系統提煉為模組化設計空間（編碼、儲存、檢索、管理），既提供標準化實現基底，也構建了公平的實驗場域。在四項具挑戰性的智能體基準測試中，廣泛實驗表明MemEvolve實現了：（I）顯著性能提升，將SmolAgent、Flash-Searcher等框架的表現最高提升17.06%；（II）強大的跨任務與跨LLM泛化能力，其設計的記憶架構能有效遷移至不同基準測試與骨幹模型。
English
Self-evolving memory systems are unprecedentedly reshaping the evolutionary paradigm of large language model (LLM)-based agents. Prior work has predominantly relied on manually engineered memory architectures to store trajectories, distill experience, and synthesize reusable tools, enabling agents to evolve on the fly within environment interactions. However, this paradigm is fundamentally constrained by the staticity of the memory system itself: while memory facilitates agent-level evolving, the underlying memory architecture cannot be meta-adapted to diverse task contexts. To address this gap, we propose MemEvolve, a meta-evolutionary framework that jointly evolves agents' experiential knowledge and their memory architecture, allowing agent systems not only to accumulate experience but also to progressively refine how they learn from it. To ground MemEvolve in prior research and foster openness in future self-evolving systems, we introduce EvolveLab, a unified self-evolving memory codebase that distills twelve representative memory systems into a modular design space (encode, store, retrieve, manage), providing both a standardized implementation substrate and a fair experimental arena. Extensive evaluations on four challenging agentic benchmarks demonstrate that MemEvolve achieves (I) substantial performance gains, improving frameworks such as SmolAgent and Flash-Searcher by up to 17.06%; and (II) strong cross-task and cross-LLM generalization, designing memory architectures that transfer effectively across diverse benchmarks and backbone models.