研究大模型在信息冲突下如何选择信息来源，发现重复能反转偏好。
Whose Facts Win? LLM Source Preferences under Knowledge Conflicts
- 用合成来源控制实验，测试模型对不同信息源的偏好。
- 模型更信任政府/报纸等机构来源，但重复可逆转此偏好。
- 提出新方法降低重复偏差79.2%，适合可信度研究者参考。
随着大语言模型（LLMs）在检索增强生成流程中的广泛应用，其在知识冲突下的行为愈发重要。然而，检索信息来源的作用尚未被系统研究。本文提出一种新框架，基于跨学科可信度研究，探究英文语境下源偏好如何影响LLM解决上下文知识冲突。通过使用合成来源，避免真实来源偏见，对13个开源权重的LLM进行严格控制评估。结果表明，模型倾向于采纳机构证实的信息（如政府或新闻来源），而非个人或社交媒体信息。然而，仅通过重复低可信度来源的内容，即可反转这一偏好。为缓解重复效应并保持稳定偏好，本文提出一种新方法，可将重复偏差降低高达79.2%，同时保留至少72.5%的原始偏好。所有数据与代码已公开，以推动知识密集型NLP中可信度与源偏好研究。
原文摘要 · Abstract (English)
As large language models (LLMs) are more frequently used in retrieval-augmented generation pipelines, it is increasingly relevant to study their behavior under knowledge conflicts. Thus far, the role of the source of the retrieved information has gone unexamined. We address this gap with a novel framework to investigate how source preferences affect LLM resolution of inter-context knowledge conflicts in English, motivated by interdisciplinary research on credibility. By using synthetic sources, we study preferences for different types of sources without inheriting the biases of specific real-world sources. With a comprehensive, tightly-controlled evaluation of 13 open-weight LLMs, we find that LLMs prefer institutionally-corroborated information (e.g., government or newspaper sources) over information from people and social media. However, these source preferences can be reversed by simply repeating information from less credible sources. To mitigate repetition effects and maintain consistent preferences, we propose a novel method that reduces repetition bias by up to 79.2%, while also maintaining at least 72.5% of original preferences. We release all data and code to encourage future work on credibility and source preferences in knowledge-intensive NLP.
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。