Towards Fine-Grained Citation Evaluation in Generated Text: A Comparative Analysis of Faithfulness Metrics

TL;DR

This paper evaluates the effectiveness of faithfulness metrics in distinguishing different levels of citation support in generated text, highlighting the complexity of fine-grained assessment and providing recommendations for improvement.

Contribution

It introduces a comprehensive evaluation framework for faithfulness metrics in fine-grained citation support scenarios, revealing their varied effectiveness.

Findings

No single metric consistently outperforms others across all evaluations

Current metrics struggle with fine-grained differentiation of citation support levels

Practical recommendations are provided for developing better faithfulness metrics

Abstract

Large language models (LLMs) often produce unsupported or unverifiable content, known as "hallucinations." To mitigate this, retrieval-augmented LLMs incorporate citations, grounding the content in verifiable sources. Despite such developments, manually assessing how well a citation supports the associated statement remains a major challenge. Previous studies use faithfulness metrics to estimate citation support automatically but are limited to binary classification, overlooking fine-grained citation support in practical scenarios. To investigate the effectiveness of faithfulness metrics in fine-grained scenarios, we propose a comparative evaluation framework that assesses the metric effectiveness in distinguishing citations between three-category support levels: full, partial, and no support. Our framework employs correlation analysis, classification evaluation, and retrieval evaluation to measure the alignment between metric scores and human judgments comprehensively. Our results show no single metric consistently excels across all evaluations, revealing the complexity of assessing fine-grained support. Based on the findings, we provide practical recommendations for developing more effective metrics.