从计算到裁决：考察数学推理任务上的 LLM 评判者
From Calculation to Adjudication: Examining LLM judges on Mathematical Reasoning Tasks
AI总结：
本文研究数学推理任务中的 LLM 裁判，揭示其评判难度与候选模型性能相关，并可用简单词性特征预测多数评判，但其难以用于直接提升任务表现。
摘要：
为了减少对人工标注的需求，大型语言模型（LLM）已被提议用作评判其他候选模型质量的裁判。LLM 裁判的性能通常通过衡量其与摘要或机器翻译等生成任务上的人类判断的相关性来评估。相比之下，我们研究数学推理任务上的 LLM 裁判。这些任务需要多步推理，且其解答的正确性是可验证的，从而能够进行更客观的评估。我们进行了详细的性能分析，发现简单样本容易评判，而困难样本难以评判。我们的分析揭示了评判性能与候选模型任务性能之间存在强相关性，表明裁判倾向于偏好更高质量的模型，即使它们的答案是错误的。因此，我们测试了能否使用词性标注等简单特征预测 LLM 裁判的行为，并发现我们能够正确预测 70%–75% 的评判结果。我们通过分析实际用例结束本研究，结果表明 LLM 裁判能够持续识别平均而言更好的模型，但如果我们用它们来提升任务性能，则会在很大程度上失败。
英文摘要：
To reduce the need for human annotations, large language models (LLMs) have been proposed as judges of the quality of other candidate models. The performance of LLM judges is typically evaluated by measuring the correlation with human judgments on generative tasks such as summarization or machine translation. In contrast, we study LLM judges on mathematical reasoning tasks. These tasks require multi-step reasoning, and the correctness of their solutions is verifiable, enabling a more objective evaluation. We perform a detailed performance analysis and find that easy samples are easy to judge, and difficult samples are difficult to judge. Our analysis uncovers a strong correlation between judgment performance and the candidate model task performance, indicating that judges tend to favor higher-quality models even if their answer is incorrect. As a consequence, we test whether we can predict the behavior of LLM judges using simple features such as part-of-speech tags and find that we can correctly predict 70%-75% of judgments. We conclude this study by analyzing practical use cases, showing that LLM judges consistently detect the on-average better model but largely fail if we use them to improve task performance.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
