AI 中文总结
针对LLM智能体故障修复的决策缺口，提出DAAF框架，通过组件属性归因学习干预效果，无需反事实重放，在电信任务上实现高命中率并提升任务成功率。
AI 中文摘要
部署的LLM智能体越来越依赖持久化、版本化的系统资产，如路由规则、知识片段、提示指令和可复用技能。故障定位方法可以识别错误在智能体或执行轨迹中的显现位置，但修复需要不同的决策：应更改哪个可编辑系统资产，且该更改是否预期能改善任务结果？我们通过组件-属性故障归因来研究这一差距，其中诊断目标是版本化、可寻址的项，而非执行位置。我们提出检测感知归因框架（DAAF），该框架学习有效属性替换的效果，并将这种干预证据摊销到部署时诊断中。DAAF结合稀疏且嘈杂的故障信号来决定是否需要进行干预，从由可执行任务结果评估的受控重放中学习组件类型条件下的替换效果，并在具有兼容干预响应的请求间共享监督。在诊断时，DAAF仅使用观察到的执行、注册的候选和可用的故障信号；它既不需要反事实重放，也不需要任务奖励，并在证据不足时返回no_change、修复目标或未解决决策。在保留的tau^2-bench电信任务上，DAAF达到80.72%的属性Hit@1，恢复了62.65%的失败执行，同时将干净任务回归限制在3.23%，并达到71.93%的整体任务成功率。这些结果表明，基于干预的属性归因可以将故障定位连接到可执行的系统修复。
英文摘要
Deployed LLM agents increasingly rely on persistent, versioned system assets such as routing rules, knowledge segments, prompt instructions, and reusable skills. Failure-localization methods can identify where an error manifests in an agent or execution trace, but repair requires a different decision: which editable system asset should be changed, and is that change expected to improve the task outcome? We study this gap through component-attribute failure attribution, where diagnosis targets versioned, addressable items rather than execution locations. We propose the Detection-Aware Attribution Framework (DAAF), which learns the effects of valid attribute replacements and amortizes this intervention evidence into deployment-time diagnosis. DAAF combines sparse and noisy failure signals to decide whether intervention is warranted, learns component-type-conditioned replacement effects from controlled replays evaluated by executable task outcomes, and shares supervision across requests with compatible intervention responses. At diagnosis time, DAAF uses only the observed execution, registered candidates, and available failure signals; it requires neither counterfactual replay nor task reward and returns no_change, a repair target, or an unresolved decision when evidence is insufficient. On held-out tau^2-bench Telecom tasks, DAAF achieves 80.72% attribute Hit@1, recovers 62.65% of failed executions while limiting clean-task regression to 3.23%, and reaches 71.93% overall task success. These results show that intervention-grounded attribute attribution can connect failure localization to executable system repair.
Comments21 pages, 1 figure. Preprint
