探索指令微调对大语言模型错误信息易感性的影响
Exploring the Impact of Instruction-Tuning on LLM's Susceptibility to Misinformation
AI总结：
本文研究指令微调如何影响LLM对错误信息的易感性，发现其会增强模型对用户输入的依赖，并分析提示角色、信息长度和系统警告等因素。
摘要：
指令微调增强了大语言模型（LLM）更准确遵循用户指令的能力，在提升可用性的同时减少了有害输出。然而，这一过程可能增加模型对用户输入的依赖，从而导致不加过滤地接受错误信息并产生幻觉。现有研究主要强调，LLM容易接受与其参数化知识相矛盾的外部信息，但关于指令微调对这一现象的直接影响，相关研究仍然很少。在本研究中，我们调查了指令微调对LLM错误信息易感性的影响。分析表明，当错误信息由用户呈现时，经过指令微调的LLM明显更有可能接受错误信息。与基础模型的比较显示，指令微调增加了模型对用户提供信息的依赖，使易感性从助手角色转移到用户角色。此外，我们还探讨了影响错误信息易感性的其他因素，例如提示结构中用户的角色、错误信息长度，以及系统提示中是否存在警告。我们的发现强调，需要采用系统性方法来缓解指令微调的非预期后果，并提升LLM在现实世界应用中的可靠性。
英文摘要：
Instruction-tuning enhances the ability of large language models (LLMs) to follow user instructions more accurately, improving usability while reducing harmful outputs. However, this process may increase the model's dependence on user input, potentially leading to the unfiltered acceptance of misinformation and the generation of hallucinations. Existing studies primarily highlight that LLMs are receptive to external information that contradict their parametric knowledge, but little research has been conducted on the direct impact of instruction-tuning on this phenomenon. In our study, we investigate the impact of instruction-tuning on LLM's susceptibility to misinformation. Our analysis reveals that instruction-tuned LLMs are significantly more likely to accept misinformation when it is presented by the user. A comparison with base models shows that instruction-tuning increases reliance on user-provided information, shifting susceptibility from the assistant role to the user role. Furthermore, we explore additional factors influencing misinformation susceptibility, such as the role of the user in prompt structure, misinformation length, and the presence of warnings in the system prompt. Our findings underscore the need for systematic approaches to mitigate unintended consequences of instruction-tuning and enhance the reliability of LLMs in real-world applications.
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。
