LLM事实核查存在认知偏差，影响证据检索结果。
(Fact) Check Your Bias
- 用Llama 3.1直接验证事实，近半数判定为证据不足。
- 不同提示下检索到的证据约50%各不相同，存在视角偏差。
- 尽管证据差异大，最终判断结果仍保持稳定，适合可信度研究。
自动事实核查系统日益依赖大语言模型（LLMs）。本文研究参数化知识偏差如何影响HerO系统（FEVER-25基准）的事实核查结果。考察了两种偏差：(1) Llama 3.1模型内部的参数知识偏差；(2) 人为注入的偏差。当直接提示执行事实核查时，Llama 3.1将近一半声明标记为“证据不足”。仅依靠其参数知识，模型能对剩余一半声明做出判断。在第二项实验中，通过提示生成支持、反驳或中立的核查文档，提示显著影响检索结果，约50%的检索证据因视角不同而唯一。值得注意的是，模型有时拒绝为它认为错误的声明生成支持性文档，产生固有负面偏见。尽管检索证据存在差异，最终判断预测在不同提示策略下表现出稳定性。代码已公开于：https://github.com/eibakke/FEVER-8-Shared-Task
原文摘要 · Abstract (English)
Automatic fact verification systems increasingly rely on large language models (LLMs). We investigate how parametric knowledge biases in these models affect fact-checking outcomes of the HerO system (baseline for FEVER-25). We examine how the system is affected by: (1) potential bias in Llama 3.1's parametric knowledge and (2) intentionally injected bias. When prompted directly to perform fact-verification, Llama 3.1 labels nearly half the claims as "Not Enough Evidence". Using only its parametric knowledge it is able to reach a verdict on the remaining half of the claims. In the second experiment, we prompt the model to generate supporting, refuting, or neutral fact-checking documents. These prompts significantly influence retrieval outcomes, with approximately 50\% of retrieved evidence being unique to each perspective. Notably, the model sometimes refuses to generate supporting documents for claims it believes to be false, creating an inherent negative bias. Despite differences in retrieved evidence, final verdict predictions show stability across prompting strategies. The code is available at: https://github.com/eibakke/FEVER-8-Shared-Task
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。