CPR: Mitigating Large Language Model Hallucinations with Curative Prompt Refinement
This addresses the issue of unreliable outputs for users of large language models, though it is incremental as it builds on existing prompt engineering methods.
The paper tackles the problem of large language model hallucinations caused by poorly structured prompts by introducing Curative Prompt Refinement (CPR), a framework that refines prompts to align user intentions, resulting in over a 90% win rate in generation quality compared to original prompts.
Recent advancements in large language models (LLMs) highlight their fluency in generating responses to diverse prompts. However, these models sometimes generate plausible yet incorrect ``hallucinated" facts, undermining trust. A frequent but often overlooked cause of such errors is the use of poorly structured or vague prompts by users, leading LLMs to base responses on assumed rather than actual intentions. To mitigate hallucinations induced by these ill-formed prompts, we introduce Curative Prompt Refinement (CPR), a plug-and-play framework for curative prompt refinement that 1) cleans ill-formed prompts, and 2) generates additional informative task descriptions to align the intention of the user and the prompt using a fine-tuned small language model. When applied to language models, we discover that CPR significantly increases the quality of generation while also mitigating hallucination. Empirical studies show that prompts with CPR applied achieves over a 90\% win rate over the original prompts without any external knowledge.