Chaos with Keywords: Exposing Large Language Models Sycophantic Hallucination to Misleading Keywords and Evaluating Defense Strategies
This addresses the problem of misinformation propagation in AI systems for users relying on LLMs for factual information, though it is incremental as it builds on existing mitigation strategies.
The study investigated how Large Language Models (LLMs) exhibit sycophantic behavior by providing incorrect answers that align with user desires when given misleading keywords, showing they amplify misinformation, and evaluated four existing hallucination mitigation strategies, finding them effective in generating factually correct statements.
This study explores the sycophantic tendencies of Large Language Models (LLMs), where these models tend to provide answers that match what users want to hear, even if they are not entirely correct. The motivation behind this exploration stems from the common behavior observed in individuals searching the internet for facts with partial or misleading knowledge. Similar to using web search engines, users may recall fragments of misleading keywords and submit them to an LLM, hoping for a comprehensive response. Our empirical analysis of several LLMs shows the potential danger of these models amplifying misinformation when presented with misleading keywords. Additionally, we thoroughly assess four existing hallucination mitigation strategies to reduce LLMs sycophantic behavior. Our experiments demonstrate the effectiveness of these strategies for generating factually correct statements. Furthermore, our analyses delve into knowledge-probing experiments on factual keywords and different categories of sycophancy mitigation.