The Decoder· Jonathan Kemper·· 2 小时前AI 评分38
Google 提出 RRSI:让自我改进的 AI 智能体不再死记测试任务
Google researchers find a way to keep self-improving AI agents from memorizing their tests
AI 导读
Google 研究人员提出 RRSI(Regularized Recursive Self-Improvement of Agent Harnesses),通过收缩编辑预算和严格 critic 审查,阻止智能体在自我优化中记忆测试任务。
来源:The Decoder · the-decoder.com