Google 研究新方法防止自我改进 AI agent 死记测试任务
原标题：Google researchers find a way to keep self-improving AI agents from memorizing their tests
AI 摘要
Google Cloud AI Research 与多所大学提出 RRSI 方法，用于防止自我改进的 AI agent 在优化自身 harness 时过拟合测试任务。RRSI 通过逐步收紧编辑预算、引入严格 critic 过滤基准特定技巧，并只接受有性能收益的算力增加，从而提升泛化能力。在八个基准上，RRSI 训练任务最高提升 14.1 分，未见任务最高提升 4.7 分，运行时 token 减少约 30%，且性能未低于基线。
正文节选
Google researchers find a way to keep self-improving AI agents from memorizing their tests AI agents that keep optimizing their own working environment quickly tend to overspecialize on their test tasks. A new method from Google Cloud AI Research and several universities aims to prevent that while also cutting compute costs. Modern AI agents wrap a fixed language model in a so-called harness, a framework of prompts, workflows, tools, memory, and logic that controls what the model sees at each st