The Decoder· Jonathan Kemper·· 2 天前AI 评分57
Google 研究人员提出 RRSI,减少智能体 harness 对测试任务的记忆
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),通过约束智能体 harness 的自我优化,减少其记忆测试任务的问题。
来源:The Decoder · the-decoder.com