探索并缓解大语言模型中的谄媚幻觉
Exploring and Mitigating Fawning Hallucinations in Large Language Models
AI总结：
针对大语言模型因对齐欺骗/误导性提示产生的谄媚幻觉问题，本研究设计两种输入诱发范式，提出无需额外训练的协作对比解码（CCD）方法，经多任务实验验证可有效缓解谄媚幻觉、提升生成内容事实性。
摘要：
大语言模型（LLM）已展现出卓越的语言理解能力。然而，当LLM使其输出与欺骗性和/或误导性提示对齐时，生成的回复可能会偏离事实信息。这类现象被称为谄媚幻觉，即模型优先与输入隐含的视角对齐，而非准确性和真实性。本研究分析了多种自然语言处理任务中的谄媚幻觉，并针对性地调整了所谓的对比解码方法以缓解谄媚幻觉。具体而言，我们设计了两种范式来生成对应的欺骗性和/或误导性输入，以稳定地诱发谄媚幻觉。随后，我们提出了协作对比解码（CCD）来处理LLM在不同任务中的谄媚幻觉问题。通过对比诱发输入与转换后的中性输入在输出分布上的偏差，所提出的CCD无需额外训练即可减少模型对欺骗性和/或误导性信息的依赖。大量实验表明，所提出的CCD能在多种任务中有效缓解谄媚幻觉，提升生成回复的事实性。
英文摘要：
Large language models (LLMs) have demonstrated exceptional proficiency in language understanding. However, when LLMs align their outputs with deceptive and/or misleading prompts, the generated responses could deviate from the de facto information. Such observations are known as fawning hallucinations, where the model prioritizes alignment with the input's implied perspective over accuracy and truthfulness. In this work, we analyze fawning hallucinations in various natural language processing tasks and tailor the so-termed contrastive decoding method for fawning-hallucination mitigation. Specifically, we design two paradigms to generate corresponding deceptive and/or misleading inputs for the consistent fawning hallucinations induction. Then, we propose the collaborative contrastive decoding (CCD) to handle the fawning hallucinations across different tasks in LLMs. By contrasting the deviation in output distribution between induced and transformed neutral inputs, the proposed CCD can reduce reliance on deceptive and/or misleading information without requiring additional training. Extensive experiments demonstrate that the proposed CCD can effectively mitigate fawning hallucinations and improve the factuality of the generated responses over various tasks.
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。
