跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.14581 · v1

When Evaluators of AI Agents Disagree: An Empirical Study of Native Scoring, LLM Judges, and Commit-Boundary Checks on tau2-bench

首次提交 2026-11-10 13:34
版本更新 2026-11-10 13:34
雷达观测 2026-11-13 10:10
arXiv ↗PDF ↗

摘要

Agentic workflows, in which a large language model (LLM) calls tools to read and write data, are evaluated in two ways. Benchmarks compare the final state against a gold answer. Deployed services have none; an LLM judge reads the transcript and decides. Practice treats the two as interchangeable. This paper measures whether they are. We ran 2,160 episodes in tau2-bench's airline and retail domains and recorded every database write. We scored each episode four ways: native scoring against gold state, an LLM judge, that judge given the policy document, and the boundary evaluator, a minimal rule-based check of whether a write was expected against whether one occurred in the trace. Two annotators independently labeled 120 episodes for task correctness, policy compliance, and commitment correctness. The judge repeated its verdict 98.2% of the time but agreed with native scoring on 56.2%. In 11 episodes the agent executed a forbidden write. The judge passed 10 to 11, with or without the policy or transcript. A stronger judge still passed 6. The boundary evaluator, with no model call, caught all 11; no judge significantly exceeded it on human labels. Forbidden writes occurred in 11.1% of episodes without gates. Verification, correction, and safe-termination gates did not reduce them significantly. The traces show a candidate cause: the gates' evidence digest dropped the order status and customer authentication, so valid requests were rejected. The choice of evaluator decides the conclusion; the minimal check exposes failures the judges miss. Traces, code, prompts, and labels are released.

作者备注:38 pages, 3 figures, 11 tables. Submitted to Empirical Software Engineering. Replication package: https://doi.org/10.5281/zenodo.23683806. Supplementary material (Online Resource 1) is included as an ancillary file

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v1 · 2026-11-10 13:34 · 最新观测版本

When Evaluators of AI Agents Disagree: An Empirical Study of Native Scoring, LLM Judges, and Commit-Boundary Checks on tau2-bench

Agentic workflows, in which a large language model (LLM) calls tools to read and write data, are evaluated in two ways. Benchmarks compare the final state against a gold answer. Deployed services have none; an LLM judge reads the transcript and decides. Practice treats the two as interchangeable. This paper measures whether they are. We ran 2,160 episodes in tau2-bench's airline and retail domains and recorded every database write. We scored each episode four ways: native scoring against gold state, an LLM judge, that judge given the policy document, and the boundary evaluator, a minimal rule-based check of whether a write was expected against whether one occurred in the trace. Two annotators independently labeled 120 episodes for task correctness, policy compliance, and commitment correctness. The judge repeated its verdict 98.2% of the time but agreed with native scoring on 56.2%. In 11 episodes the agent executed a forbidden write. The judge passed 10 to 11, with or without the policy or transcript. A stronger judge still passed 6. The boundary evaluator, with no model call, caught all 11; no judge significantly exceeded it on human labels. Forbidden writes occurred in 11.1% of episodes without gates. Verification, correction, and safe-termination gates did not reduce them significantly. The traces show a candidate cause: the gates' evidence digest dropped the order status and customer authentication, so valid requests were rejected. The choice of evaluator decides the conclusion; the minimal check exposes failures the judges miss. Traces, code, prompts, and labels are released.

查看此版本解读 →查看此版本原文 ↗