Improving Factuality and Reasoning in Language Models through Multiagent Debate

TL;DR

This paper introduces a multi-agent debate framework where multiple language models argue and reason over multiple rounds, significantly improving factual accuracy and reasoning in language model outputs.

Contribution

The paper presents a novel multi-agent debate approach that enhances reasoning and factuality in language models without modifying their core architecture.

Findings

Improves mathematical and strategic reasoning performance.

Reduces hallucinations and fallacious answers.

Applicable to existing black-box models without retraining.

Abstract

Large language models (LLMs) have demonstrated remarkable capabilities in language generation, understanding, and few-shot learning in recent years. An extensive body of work has explored how their performance may be further improved through the tools of prompting, ranging from verification, self-consistency, or intermediate scratchpads. In this paper, we present a complementary approach to improve language responses where multiple language model instances propose and debate their individual responses and reasoning processes over multiple rounds to arrive at a common final answer. Our findings indicate that this approach significantly enhances mathematical and strategic reasoning across a number of tasks. We also demonstrate that our approach improves the factual validity of generated content, reducing fallacious answers and hallucinations that contemporary models are prone to. Our approach may be directly applied to existing black-box models and uses identical procedure and prompts for all tasks we investigate. Overall, our findings suggest that such "society of minds" approach has the potential to significantly advance the capabilities of LLMs and pave the way for further breakthroughs in language generation and understanding.