FactLens: Benchmarking Fine-Grained Fact Verification

TL;DR

FactLens introduces a benchmark for fine-grained fact verification that breaks down complex claims into sub-claims, enabling more precise error detection and transparency in verifying LLM outputs.

Contribution

This paper presents FactLens, a novel benchmark with metrics and evaluators for assessing the quality of sub-claims in fine-grained fact verification tasks.

Findings

Automated evaluators align well with human judgments.

Sub-claim characteristics significantly affect verification performance.

High-quality, manually curated ground truth data enhances benchmark reliability.

Abstract

Large Language Models (LLMs) have shown impressive capability in language generation and understanding, but their tendency to hallucinate and produce factually incorrect information remains a key limitation. To verify LLM-generated contents and claims from other sources, traditional verification approaches often rely on holistic models that assign a single factuality label to complex claims, potentially obscuring nuanced errors. In this paper, we advocate for a shift towards fine-grained verification, where complex claims are broken down into smaller sub-claims for individual verification, allowing for more precise identification of inaccuracies, improved transparency, and reduced ambiguity in evidence retrieval. However, generating sub-claims poses challenges, such as maintaining context and ensuring semantic equivalence with respect to the original claim. We introduce FactLens, a benchmark for evaluating fine-grained fact verification, with metrics and automated evaluators of sub-claim quality. The benchmark data is manually curated to ensure high-quality ground truth. Our results show alignment between automated FactLens evaluators and human judgments, and we discuss the impact of sub-claim characteristics on the overall verification performance.