"merge Conflicts!" Exploring The Impacts Of External Distractors To Parametric Knowledge Graphs

Qian Cheng, Zhao Xinran, Wu Sherry Tongshuang. Arxiv 2023

[Paper]    
Applications Reinforcement Learning Tools Training Techniques

Large language models (LLMs) acquire extensive knowledge during pre-training, known as their parametric knowledge. However, in order to remain up-to-date and align with human instructions, LLMs inevitably require external knowledge during their interactions with users. This raises a crucial question: How will LLMs respond when external knowledge interferes with their parametric knowledge? To investigate this question, we propose a framework that systematically elicits LLM parametric knowledge and introduces external knowledge. Specifically, we uncover the impacts by constructing a parametric knowledge graph to reveal the different knowledge structures of LLMs, and introduce external knowledge through distractors of varying degrees, methods, positions, and formats. Our experiments on both black-box and open-source models demonstrate that LLMs tend to produce responses that deviate from their parametric knowledge, particularly when they encounter direct conflicts or confounding changes of information within detailed contexts. We also find that while LLMs are sensitive to the veracity of external knowledge, they can still be distracted by unrelated information. These findings highlight the risk of hallucination when integrating external knowledge, even indirectly, during interactions with current LLMs. All the data and results are publicly available.

Similar Work