MiniCheck: Efficient Fact-Checking of LLMs on Grounding Documents

arXiv:2404.10774v1 Announce Type: new
Abstract: Recognizing if LLM output can be grounded in evidence is central to many tasks in NLP: retrieval-augmented generation, summarization, document-grounded dialogue, and more. Current approaches to this kind of "fact-checking" are based on verifying each piece of a model generation against potential evidence using an LLM. However, this process can be very computationally expensive, requiring many calls to LLMs to check a single response. In this work, we show how to build small models that …

abstract arxiv cs.ai cs.cl current dialogue document documents evidence fact-checking however kind llm llms nlp retrieval retrieval-augmented summarization tasks type