AI 中文总结
本研究在受控文件恢复环境中发现，不考虑成本的监管者会因干预强度增加而引发工具故障并阻塞任务，而基于已知诱发事件移动平均的自适应退避规则可降低干预概率、减少阻塞并提升完成率，实验验证了其有效性。
AI 中文摘要
监督管理者可能干扰其所监管的工具使用智能体。我们在一个受控的文件恢复环境中研究这种可能性，在该环境中，监管强度的增加会触发实验性强加的工具故障。一个不考虑成本的监管者可能将这些故障转变为持续阻塞，从而阻止任务完成。我们将该监管者与一种退避规则进行比较，该规则利用已知诱发事件的移动平均值来降低干预概率。在一个手工编码的随机策略智能体上，故障模式在结果替换和执行损坏的工具参数两种情况下均出现。对于持续策略，自适应退避相比干预频率大致匹配的固定弱监管者，提高了任务完成率。一个包含6个任务、共576个回合的Gemini 2.5 Flash实验也显示，在退避机制下阻塞减少且完成率提高；在测试的设置中，中等强度的退避取得了最高的总体成功次数。这些结果揭示了干预成本与持续动作阻塞之间的相互作用，并提供了一种可能的缓解措施。成本机制是强加的，其诱发事件对退避规则直接可观察；该机制在受控环境之外的适用性仍是一个实证问题。
英文摘要
Supervisory governors can interfere with the tool-using agents they regulate. We study this possibility in a controlled file-recovery environment where increases in regulatory intensity trigger experimentally imposed tool failures. A cost-blind governor can turn these failures into persistent blocking that prevents task completion. We compare this governor with a backoff rule that reduces intervention probability using a moving average of known induced events. On a hand-coded stochastic-policy agent, the failure pattern appears under both result replacement and execution of corrupted tool arguments. For the persistent policy, adaptive backoff improves completion relative to a fixed weak governor with approximately matched intervention frequency. A Gemini 2.5 Flash experiment comprising 576 episodes across 6 tasks also shows reduced blocking and improved completion under backoff; among the tested settings, intermediate backoff strength achieves the highest observed aggregate success. These results identify an interaction between intervention cost and persistent action blocking, together with a possible mitigation. The cost mechanisms are imposed and their induced events are directly observable to the backoff rule; applicability beyond this controlled environment remains an empirical question.
CommentsIncludes ancillary experimental data and a Python script for reproducing numerical summaries. Code: https://github.com/CriticalAttentionSystems/CASAgentGovernor
