Genetic Approach to Mitigate Hallucination in Generative IR

TL;DR

This paper introduces a genetic algorithm with a balanced fitness function to significantly reduce hallucinations in generative information retrieval, improving factual accuracy while maintaining relevance.

Contribution

It proposes a novel balanced fitness function combining relevance and grounding metrics, enhancing factual accuracy in generative IR models.

Findings

Quadruples grounded answer accuracy

Maintains high relevance in generated answers

Effective mitigation of hallucinations

Abstract

Generative language models hallucinate. That is, at times, they generate factually flawed responses. These inaccuracies are particularly insidious because the responses are fluent and well-articulated. We focus on the task of Grounded Answer Generation (part of Generative IR), which aims to produce direct answers to a user's question based on results retrieved from a search engine. We address hallucination by adapting an existing genetic generation approach with a new 'balanced fitness function' consisting of a cross-encoder model for relevance and an n-gram overlap metric to promote grounding. Our balanced fitness function approach quadruples the grounded answer generation accuracy while maintaining high relevance.