The Decoder· Jonathan Kemper·· 3 小时前AI 评分63
Google 提出 RRSI 防止自优化智能体背熟测试任务
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),用递减的编辑预算和批评者过滤基准专属技巧,抑制智能体递归自优化对测试任务的记忆。
来源:The Decoder · the-decoder.com