Understanding the Impact of Confidence in Retrieval Augmented Generation: A Case Study in the Medical Domain

TL;DR

This study investigates how Retrieval Augmented Generation (RAG) influences the confidence levels of Large Language Models in the medical domain, analyzing whether models can assess the relevance of retrieved documents through output probabilities.

Contribution

It provides an empirical analysis of confidence mechanisms in RAG models within the medical field, highlighting their ability to judge document relevance based on output probabilities.

Findings

Certain models can assess the relevance of retrieved documents.

Model confidence correlates with answer correctness.

Evaluation metrics like calibration error and entropy are effective.

Abstract

Retrieval Augmented Generation (RAG) complements the knowledge of Large Language Models (LLMs) by leveraging external information to enhance response accuracy for queries. This approach is widely applied in several fields by taking its advantage of injecting the most up-to-date information, and researchers are focusing on understanding and improving this aspect to unlock the full potential of RAG in such high-stakes applications. However, despite the potential of RAG to address these needs, the mechanisms behind the confidence levels of its outputs remain underexplored. Our study focuses on the impact of RAG, specifically examining whether RAG improves the confidence of LLM outputs in the medical domain. We conduct this analysis across various configurations and models. We evaluate confidence by treating the model's predicted probability as its output and calculating several evaluation metrics which include calibration error method, entropy, the best probability, and accuracy. Experimental results across multiple datasets confirmed that certain models possess the capability to judge for themselves whether an inserted document relates to the correct answer. These results suggest that evaluating models based on their output probabilities determine whether they function as generators in the RAG framework. Our approach allows us to evaluate whether the models handle retrieved documents.