Evaluating the Effectiveness of Retrieval-Augmented Large Language Models in Scientific Document Reasoning
This highlights a critical reliability issue for researchers and practitioners using AI in scientific contexts, though it is incremental as it builds on known hallucination problems.
The study evaluated retrieval-augmented large language models on scientific document reasoning tasks and found that these models often justify predictions with fabricated evidence, even when pretrained on scientific data.
Despite the dramatic progress in Large Language Model (LLM) development, LLMs often provide seemingly plausible but not factual information, often referred to as hallucinations. Retrieval-augmented LLMs provide a non-parametric approach to solve these issues by retrieving relevant information from external data sources and augment the training process. These models help to trace evidence from an externally provided knowledge base allowing the model predictions to be better interpreted and verified. In this work, we critically evaluate these models in their ability to perform in scientific document reasoning tasks. To this end, we tuned multiple such model variants with science-focused instructions and evaluated them on a scientific document reasoning benchmark for the usefulness of the retrieved document passages. Our findings suggest that models justify predictions in science tasks with fabricated evidence and leveraging scientific corpus as pretraining data does not alleviate the risk of evidence fabrication.