MetaShare：面向开放式智能体的失败感知集体自改进
原标题：MetaShare: Failure-Aware Collective Self-Improvement for Open-Ended Agents
AI 摘要
论文提出 MetaShare，一个面向开放式自改进智能体的失败感知集体自改进框架。它将每次自修改记录为带符号性能增量的机制级记录，成功与失败分通道检索，并由任务盲审计器补充缺失机制类别、基于嵌入去重。在论文评审、编程、数学评分、机器人奖励设计和金融审计风险五个领域，MetaShare 的中位表现匹配或超过 DGM-Hyperagents，并少用 1.7–2.9 次进化迭代达到自身最佳智能体。
正文节选
Managing Agents that Manage Agents: Workshop on Responsible Use of Meta-Agents MetaShare: Failure-Aware Collective Self-Improvement for Open-Ended Agents Abstract Humans often learn fastest not by repeating success, but by remembering which failures not to repeat. Open-ended self-improving agents can share experience across lineages, as in GEA, or edit their own improvement mechanisms, as in DGM-Hyperagents, yet failed improvement attempts are rarely retained as reusable cross-branch knowledge,