跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2606.07379 · v3

Do Coding Agents Deceive Us? Detecting and Preventing Cheating via Capped Evaluation with Randomized Tests

首次提交 2026-06-05 23:20
版本更新 2026-11-16 16:37
雷达观测 2026-11-17 17:11
arXiv ↗PDF ↗

摘要

A growing failure mode in agent evaluation and training is that models can achieve high evaluation scores (e.g., test pass rates) by exploiting shortcuts instead of solving the intended task, making these scores deceptive. This makes evaluation scores unreliable as measures of true task-solving ability. We propose CapCode, a framework for constructing coding datasets with randomized tests whose best achievable non-cheating score is deliberately capped below 100%. This capped-score design gives evaluation scores a clearer interpretation: scores substantially above the cap are implausible and therefore provide evidence of cheating. To prevent cheating, we propose CapReward, a reward design based on the CapCode principle to discourage optimization beyond the cap. Experiments across multiple datasets show that CapCode detects cheating while preserving performance ranking of models, and CapReward reduces cheating behavior, yielding models that better follow the intended task specification.

作者备注:NeurIPS 2026 camera ready version

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

本站首次收录为 v3,暂无更早版本可对照。

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。

v3 · 2026-11-16 16:37 · 最新观测版本

Do Coding Agents Deceive Us? Detecting and Preventing Cheating via Capped Evaluation with Randomized Tests

A growing failure mode in agent evaluation and training is that models can achieve high evaluation scores (e.g., test pass rates) by exploiting shortcuts instead of solving the intended task, making these scores deceptive. This makes evaluation scores unreliable as measures of true task-solving ability. We propose CapCode, a framework for constructing coding datasets with randomized tests whose best achievable non-cheating score is deliberately capped below 100%. This capped-score design gives evaluation scores a clearer interpretation: scores substantially above the cap are implausible and therefore provide evidence of cheating. To prevent cheating, we propose CapReward, a reward design based on the CapCode principle to discourage optimization beyond the cap. Experiments across multiple datasets show that CapCode detects cheating while preserving performance ranking of models, and CapReward reduces cheating behavior, yielding models that better follow the intended task specification.

查看此版本解读 →查看此版本原文 ↗