Newest papers

What the Trace Shows and What the Label Says: Auditing Behavioral Detectors in Agent Evaluation

Idunnuoluwa A. Adeniji, Shilpika, Michael E. Papka · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.30491 · cited by 0 works (OpenAlex) · open-access PDF

Japanese

これらのラベルは、パッサー、行動検出器、およびハーネスルールによって生成された測定値の明確な真実を示すと、これらのラベルは、パッサー、行動検出器、およびハーネスルールによって生成された測定値です。私たちは6つの生産モデルおよび30の問題を完全に収集しません。最終正確性は、GSM8Kに構築されたツール使用設定で、エージェントは、出力の証拠が時々破損する可能性がある計算機を呼び出します。これらのラベルは、ツールが返したもの、モデルが後で使用したもの、および検出器が報告したものについて明確な真実を示します。私たちは、6つの生産モデルおよび30のリムシ

Machine translation (Meta M2M100); the English original follows.

Abstract

LLM agents are often evaluated by reducing execution traces to labels such as final-answer correctness, tool verification, and fault propagation. These labels are measurements produced by parsers, behavioral detectors, and harness rules, and they may not fully capture what the trace shows. We study this in a tool-use setting built on GSM8K, where agents call a calculator whose outputs are occasionally corrupted, giving clear ground truth for what the tool returned, what the model later used, and what the detectors reported. We audit three trace-level detectors across six production models and 30 problems per model. Final correctness, visible re-verification, and fault propagation can disagree in systematic ways. A key failure mode is silent tool-output override: the calculator returns a corrupted value, but the model later uses a different, often correct, value without calling the tool again, so the bad value stops propagating while a verification detector records no re-check. Human review of 90 traces agrees perfectly with the automated labels for final correctness (67/67), fault impact (19/19), and fault relevance (18/18), but only 67/81 for verification; all 14 disagreements fall into four recurring patterns. Execution limits also change the reported result: in an ablation across two models, one ranges from 21/30 to 27/30 correct as step and token budgets change, while the other stays between 27/30 and 29/30. These findings motivate measurement transparency: an automated label should come with the trace evidence behind it, whether the question the detector asks applied to that trace, whether the run was cut off by an execution limit, and any human corrections. Our trace viewer makes this information inspectable alongside the reported results.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: TRACE (psycholinguistics), Correctness, Calculator, Computer science, Detector, Security token

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange