智能体步骤值：使用基于状态的语言模型评估器进行状态转换测量
Agent Step Value: Auditing Evaluator-Channel Reversals in Black-Box Agent Traces
浏览论文内容
中文总结 AI 辅助
研究提出智能体步骤值（ASV）框架，通过状态转换测量评估智能体动作，能定位最终答案分数等遗漏的关键信息，还介绍了具体方法及工具，并用实验验证其有效性。
中文摘要 AI 辅助
大多数智能体评估将多步轨迹简化为最终答案、成功标志或轨迹级分数。这些汇总掩盖了开发者最需要的诊断问题。我们引入智能体步骤值（ASV），一个状态转换测量框架，通过状态评估器分布变化对动作评分，呈现编辑后的状态投影，使用无状态语言模型评估器，报告多种指标，发布独立工具包并验证其有效性。
英文摘要
Pooling, substituting, or reusing evaluator-derived step rewards assumes that their direction survives a change of evaluation channel. The same frozen transition can violate that assumption. Process rewards vary agent states, while evaluator audits vary scoring configurations; neither first difference isolates their interaction. We define Agent Step Value (ASV) as a channel-indexed target-margin gain and identify the state-by-channel interaction on complete matched faces. Across frozen PubMed question-answering transitions, direct scoring yields a positive mean ASV, while the generated-view channel yields a negative mean. Two matched replay waves reproduce this reversal, and cross-channel sign disagreement exceeds same-channel retry disagreement by 48.0 percentage points. Matched retrieval faces localize the reversal to the generated-view coordinate and trace its direction across a readout-and-stack bridge. A source-only generation contract restores the positive mean direction on artifact-bearing retrievals and removes parser-detected substantive support claims from artifact-free before-state views. ASV turns channel sensitivity into an identified measurement problem that can be localized and tested by intervention before step rewards are reused.
发表机构
- KTH Royal Institute of Technology（皇家理工学院）
- University of Electronic Science and Technology of China（电子科技大学）
机构由 AI 辅助整理，请以论文原文为准。
