过于一致而难以检测：大语言模型中自一致错误的研究
Too Consistent to Detect: A Study of Self-Consistent Errors in LLMs
AI总结：
本研究定义并检测大语言模型中的自一致错误，发现其频率随模型规模增大不降反升且现有方法难以检测，提出融合外部验证器隐藏状态的跨模型探针方法，显著提升检测性能。
摘要：
由于大语言模型（LLMs）经常生成看似合理但不正确的内容，错误检测对于确保真实性变得日益关键。然而，现有的检测方法往往忽视了一个我们称之为“自一致错误”的关键问题，即LLMs在多次随机采样中反复生成相同的错误回答。本研究正式定义了自一致错误，并评估了主流检测方法在这类错误上的表现。我们的调查揭示了两个关键发现：（1）与不一致错误不同，其频率随着LLM规模增大而显著降低，自一致错误的频率保持稳定甚至增加。（2）所有四类检测方法在检测自一致错误方面均存在显著困难。这些发现揭示了当前检测方法的关键局限性，并强调了改进的必要性。受自一致错误在不同LLM间往往表现不同的观察启发，我们提出了一种简单但有效的跨模型探针方法，该方法融合来自外部验证器LLM的隐藏状态证据。我们的方法在三个LLM家族上显著提升了自一致错误的检测性能。
英文摘要：
As large language models (LLMs) often generate plausible but incorrect content, error detection has become increasingly critical to ensure truthfulness. However, existing detection methods often overlook a critical problem we term as self-consistent error, where LLMs repeatedly generate the same incorrect response across multiple stochastic samples. This work formally defines self-consistent errors and evaluates mainstream detection methods on them. Our investigation reveals two key findings: (1) Unlike inconsistent errors, whose frequency diminishes significantly as the LLM scale increases, the frequency of self-consistent errors remains stable or even increases. (2) All four types of detection methods significantly struggle to detect self-consistent errors. These findings reveal critical limitations in current detection methods and underscore the need for improvement. Motivated by the observation that self-consistent errors often differ across LLMs, we propose a simple but effective cross-model probe method that fuses hidden state evidence from an external verifier LLM. Our method significantly enhances performance on self-consistent errors across three LLM families.
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。
