ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate

TL;DR

ChatEval introduces a multi-agent debate framework where multiple LLMs collaboratively evaluate responses, improving the reliability and human-like quality of assessments for open-ended and NLG tasks.

Contribution

This work pioneers a multi-agent debate approach with LLMs acting as collaborative evaluators, surpassing single-agent methods in response quality assessment.

Findings

ChatEval achieves more reliable evaluations than single-agent models.

Multi-agent debate enhances assessment accuracy for complex NLG tasks.

The approach mimics human evaluation processes more closely.

Abstract

Text evaluation has historically posed significant challenges, often demanding substantial labor and time cost. With the emergence of large language models (LLMs), researchers have explored LLMs' potential as alternatives for human evaluation. While these single-agent-based approaches show promise, experimental results suggest that further advancements are needed to bridge the gap between their current effectiveness and human-level evaluation quality. Recognizing that best practices of human evaluation processes often involve multiple human annotators collaborating in the evaluation, we resort to a multi-agent debate framework, moving beyond single-agent prompting strategies. The multi-agent-based approach enables a group of LLMs to synergize with an array of intelligent counterparts, harnessing their distinct capabilities and expertise to enhance efficiency and effectiveness in handling intricate tasks. In this paper, we construct a multi-agent referee team called ChatEval to autonomously discuss and evaluate the quality of generated responses from different models on open-ended questions and traditional natural language generation (NLG) tasks. Our analysis shows that ChatEval transcends mere textual scoring, offering a human-mimicking evaluation process for reliable assessments. Our code is available at https://github.com/chanchimin/ChatEval.