他们放弃的计划，他们撰写的报告：自主代理的叙事层
Plans They Abandon, Reports They Author: The Narrative Layer of Autonomous Agents
浏览论文内容
中文总结 AI 辅助
本研究基于大量真实会话数据，探究编码代理自撰摘要的覆盖度与偏向性，发现其仅反映少量操作且随执行偏离计划而更接近计划，并强调验证测量步骤的重要性。
中文摘要 AI 辅助
当编码代理完成任务时，开发者审查的是代理自己撰写的摘要，而非某人设计的展示。我们探究该摘要承载了代理工作的多少内容，以及当执行偏离既定计划时，摘要是否倾向于代理最初陈述的计划。在5,851个真实开发者会话和355,942次工具调用中，自我报告大约提及了十一分之一的操作，而仅依据报告工作的读者恢复了约五分之一的行动日志。这两个数字均不取决于会话之后是否需要人工修正。报告通常并不比已执行的操作更接近陈述的计划，但随着执行与计划的偏离加剧，报告确实越来越接近计划。我们对手动验证使用语言模型的两个测量步骤，报告了失败的一个和通过的一个，并且仅从经得起检验的测量中得出结论。
英文摘要
When a coding agent finishes a task, the developer reviews a summary the agent wrote about itself, not a display someone designed. We ask how much of the agent's work that summary carries, and whether it drifts toward the plan the agent stated when execution departed from it. Across 5,851 real developer sessions and 355,942 tool calls, a self-report referred to about one action in eleven, and a reader working from the report alone recovered roughly a fifth of the action log. Neither figure depended on whether the session later needed human correction. Reports did not generally resemble the stated plan more than the executed one, but they did so increasingly as execution diverged from the plan. We hand-validate both measurement steps that use a language model, report the one that failed alongside the one that passed, and draw conclusions only from measures that survived.
