发表机构
School of AI for Science, Peking University; School of Electronic and Computer Engineering, Peking University; School of Computer Science, Peking University（北京大学AI科学学院; 北京大学电子与计算机工程学院; 北京大学计算机学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
LabEvolver是无训练框架，通过内外部循环为湿实验室智能体配备情景记忆，在溶液制备任务中提升效率与安全性，在ALFWorld中也提升了连续任务成功率，支持闭环自动化科学发现。
AI 中文摘要
我们提出LabEvolver，这是一种无训练框架，可为安全且务实的湿实验室智能体配备来自执行经验的情景记忆。LabEvolver将基于状态的内部试验循环与外部演化循环相结合，前者用于自适应感知、在线规划和安全验证，后者将完成的轨迹提炼为可复用的技能、策略和安全经验。在机器人溶液制备任务中，LabEvolver展现出实际可行性，使pH调节完成时间减少48.2%，安全拦截次数减少60.0%。在ALFWorld中，它进一步将500个连续任务在20步内的累计成功率从ReAct方法的76.2%提升至91.4%，展现出超出湿实验室场景的通用性。这些结果表明，通过实践学习的经验演化是实现闭环自动化科学发现的可行路径，项目页面可通过指定URL访问。
英文摘要
We introduce LabEvolver, a training-free framework that equips safe and grounded wet-lab agents with episodic memory from execution experience. LabEvolver couples a state-grounded inner trial loop for adaptive perception, online planning, and safety validation with an outer evolution loop that distills completed trajectories into reusable skill, strategy, and safety experience. On robotic solution-preparation tasks, LabEvolver demonstrates real-world feasibility, reducing pH-regulation completion time and safety-gate intercepts by 48.2% and 60.0%, respectively. On ALFWorld, it further improves cumulative success rate within 20 steps from 76.2% with ReAct to 91.4% over 500 continual tasks, showing generality beyond wet-lab settings. These results support learn-by-doing experience evolution as a feasible path toward closed-loop automated scientific discovery. The project page is available at https://andygao6186.github.io/LabEvolver/.
