谁是你的评判者？论LLM生成评判的可检测性
Who's Your Judge? On the Detectability of LLM-Generated Judgments
AI总结：
本文形式化仅依据评判分数与候选内容识别LLM生成评判的任务，提出融合语言及LLM增强特征的轻量透明检测器J-Detector，实验验证其检测效果、偏见量化能力与实际效用。
摘要：
基于大语言模型（LLM）的评判利用强大的LLM高效评估候选内容并提供评判分数。然而，LLM生成评判的固有偏见和脆弱性引发了担忧，凸显出在学术同行评审等敏感场景中对其加以区分的迫切需求。在本研究中，我们提出并形式化了评判检测任务，系统研究LLM生成评判的可检测性。与LLM生成文本检测不同，评判检测仅依赖评判分数和候选内容，反映了检测过程中通常无法获得文本反馈的真实场景。我们的初步分析表明，现有LLM生成文本检测方法表现不佳，原因是它们无法捕捉评判分数与候选内容之间的交互——而这一方面对于有效评判检测至关重要。受此启发，我们提出J-Detector，这是一种轻量且透明的神经检测器，通过显式提取的语言特征和LLM增强特征进行增强，将LLM评判者的偏见与候选内容的属性关联起来，以实现准确检测。跨多个多样化数据集的实验证明了J-Detector的有效性，并展示了其可解释性如何能够量化LLM评判者中的偏见。最后，我们分析了影响LLM生成评判可检测性的关键因素，并验证了评判检测在真实场景中的实用价值。
英文摘要：
Large Language Model (LLM)-based judgments leverage powerful LLMs to efficiently evaluate candidate content and provide judgment scores. However, the inherent biases and vulnerabilities of LLM-generated judgments raise concerns, underscoring the urgent need for distinguishing them in sensitive scenarios like academic peer reviewing. In this work, we propose and formalize the task of judgment detection and systematically investigate the detectability of LLM-generated judgments. Unlike LLM-generated text detection, judgment detection relies solely on judgment scores and candidates, reflecting real-world scenarios where textual feedback is often unavailable in the detection process. Our preliminary analysis shows that existing LLM-generated text detection methods perform poorly given their incapability to capture the interaction between judgment scores and candidate content -- an aspect crucial for effective judgment detection. Inspired by this, we introduce \textit{J-Detector}, a lightweight and transparent neural detector augmented with explicitly extracted linguistic and LLM-enhanced features to link LLM judges' biases with candidates' properties for accurate detection. Experiments across diverse datasets demonstrate the effectiveness of \textit{J-Detector} and show how its interpretability enables quantifying biases in LLM judges. Finally, we analyze key factors affecting the detectability of LLM-generated judgments and validate the practical utility of judgment detection in real-world scenarios.
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。
