让智能体自动进化控制策略，持续提升长期复杂任务表现。
OpenLoopEvolve: A Verifiable Self-Evolution Framework for Loop Policies in Long-Horizon Complex Tasks
- 将控制策略封装为可追踪版本的资产，支持在线/离线演化。
- 在模拟任务中提升任务成功率与风险控制能力，优于固定初始策略。
- 适合需要长期决策与自我优化的智能体系统研究者。
长期复杂任务要求智能体在不断变化的环境中持续观察状态、制定计划、调用工具、验证结果并从失败中恢复。然而，这类控制经验常局限于单一场景或固定提示，难以积累与复用。本文提出基于环状策略（Loop Policy）的自进化框架 OpenLoopEvolve（OLE），将观察、规划、记忆、动作、验证、恢复、终止和预算控制等环节抽象为可移植的策略资产，具备版本与谱系信息。提供在线与离线两种演化模式：在线模式根据连续运行反馈生成候选策略；离线模式从历史记录与失败证据中搜索候选。两者共享由大语言模型自主提案、冠军-挑战者配对评估及稳健发布组成的演化机制。在线释放的策略在下一任务边界激活，通过后续反馈监控，若性能下降则回滚至父版本。在模拟业务基准 YC-Bench 上，两种模式均显著提升整体任务表现、成功率与风险指标，优于固定初始环状策略。结果表明，将环状策略视为可控资产，可有效积累、比较、发布与复用控制经验，从而提升智能体在长期复杂任务中的性能。
原文摘要 · Abstract (English)
Long-horizon complex tasks require agents to repeatedly observe states, formulate plans, invoke tools, verify results, and recover from failures in continuously changing environments. However, such control experience often remains confined to a single context or a fixed prompt, and is difficult to accumulate and reuse across historical traces. This paper presents OpenLoopEvolve (OLE), a self-evolution framework centered on the Loop Policy. OLE represents an agent's observation, planning, memory, action, verification, recovery, stopping, and budget control as portable policy assets with versions and lineages, and provides online and offline evolution modes that can be selected according to practical needs: the online mode triggers candidate generation based on feedback from continuous operation, whereas the offline mode searches for candidate policies from archived traces and failure evidence. Both modes share an evolution mechanism consisting of autonomous proposals by a large language model, Champion--Challenger paired evaluation, and robust release. Policies released online are activated at a subsequent task boundary, monitored using subsequent feedback, and rolled back to their parent versions when degradation conditions are met. On the simulated business benchmark YC-Bench, both modes improve aggregate task performance, task success rate, and risk metrics relative to a fixed initial Loop Policy. The results indicate that treating the Loop Policy as a governable asset can support the accumulation, comparison, release, and reuse of control experience and improve agent performance on long-horizon complex tasks.
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。