Improving Factuality and Reasoning in Language Models through Multiagent Debate

By Yilun Du, Shuang Li, Antonio Torralba, Josh Tenenbaum, Igor Mordatch. International Conference on Machine Learning, 2023

1,951 citations recorded by the source corpus.

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.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/4780d0a027c5c5a8e01d7cf697f6296880ffc945

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs (2024, positive): “Another line of work emphasizes the early stopping strategy to save computation budget while reasoning (Manvi et al., 2024; Li et al., 2024).”
Why Do Multi-Agent LLM Systems Fail? (2025, positive): “This growing interest comes from the ability of agentic systems to handle complex, multi-step tasks while dynamically interacting with diverse environments, making LLM-based agentic systems well-suited for real-world problems (Li et al., 2023).”
To CoT or not to CoT? Chain-of-thought helps mainly on math and symbolic reasoning (2024, neutral): “There is evidence that using additional calls to LLMs can help (Du et al., 2023; Yao et al., 2023; Besta et al., 2023; Chen et al., 2024), but these methods use significantly increased computation, and careful benchmarking sometimes reveals that naive techniques are as good as iterative ones…”
ChatGPT: The End of Online Exam Integrity? (2022, positive): “Even if the possibility of LLM hallucination is deemed to be likely in certain assessments, recent studies demonstrate that with careful strategies and prompting techniques, hallucination can to a large degree be mitigated, though not entirely eliminated [22,50–53].”
Large Language Model based Multi-Agents: A Survey of Progress and Challenges (2024, background): “For example, many applications [ Du et al. , 2023; Xiong et al. , 2023; Chan et al. , 2023 ] utilize multiple agents to debate a question to reach a consensus.”
Encouraging Divergent Thinking in Large Language Models through Multi-Agent Debate (2023, background): “Concurrent with our work, a few studies (Xiong et al., 2023; Du et al., 2023) also explore the multi-agent debate framework to enhance the reasoning ability of LLMs.”
Benchmarking and Defending against Indirect Prompt Injection Attacks on Large Language Models (2023, neutral): “Due to the importance of prompts, numerous works on prompt learning have been proposed [12], [55].”
Large Language Model Alignment: A Survey (2023, background): “Debate (Irving et al., 2018; Irving and Askell, 2019; Du et al., 2023) is another promising scalable oversight paradigm that can not only achieve single-agent alignment but also enable multi-agent alignment.”
A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questions (2023, neutral): “This feedback typically comes from external knowledge sources through retrieval processes (He et al., 2023; Gou et al., 2023), interactive debates with other LLMs (Du et al., 2023; Cohen et al., 2023), or guidance from external evaluation metrics (Lei et al., 2023; Khalifa et al., 2023).”
Building Cooperative Embodied Agents Modularly with Large Language Models (2023, background): “…multiple LLMs to cooperate or debate with each other "in mind" to strengthen the single agent’s capability to solve complex tasks (Li et al., 2023b; Du et al., 2023; Wang et al., 2023c), different from their "free self-talk" setting, our decentralized language agents must plan about when and what…”
Evaluating Large Language Models at Evaluating Instruction Following (2023, methodology): “Inspired by Du et al., 2023, we first prompt the LLM evaluator to give its preference using CoT with orders O 1 , O 2 and O 2 , O 1 .”
DoLa: Decoding by Contrasting Layers Improves Factuality in Large Language Models (2023, background): “Recent strategies involve inference-time self-consistency checks (Manakul et al., 2023), multi-agent debating (Du et al., 2023; Liang et al., 2023), and inference-time intervention using human labels (Li et al., 2023).”
ReConcile: Round-Table Conference Improves Reasoning via Consensus among Diverse LLMs (2023, methodology): “Across seven benchmarks spanning commonsense reasoning, mathematical reasoning, logical reasoning, and Natural Language Inference (NLI), R E C ONCILE outperforms prior single-agent (e.g., Self-Refine (Madaan et al., 2023) and Self-consistency (Wang et al., 2023b)) and multi-agent baselines (Debate (”
MedAgents: Large Language Models as Collaborators for Zero-shot Medical Reasoning (2023, background): “…witnessed the success of multi-agent collaboration (Xi et al., 2023; Wang et al., 2023d) which highlights the simulation of human activities (Du et al., 2023; Liang et al., 2023; Park et al., 2023) and optimizes the collective power of multiple agents (Chen et al., 2023; Li et al., 2023d;…”
ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate (2023, background): “Unlike previous work like Du et al. (2023), we do not explicitly ask the debater agents to reach a consensus at the end of the debate.”
Siren's Song in the AI Ocean: A Survey on Hallucination in Large Language Models (2023, neutral): “Du et al. (2023) is a pioneering work in this line.”
Multi-Agent Collaboration Mechanisms: A Survey of LLMs (2025, positive): “Researchers have found taking multiple LLM instances to debate for a fixed number of rounds can boost their factuality and reasoning capabilities [41, 77, 140].”
A survey on large language model based autonomous agents (2023, positive): “This method seeks to leverage dialogue context across multiple rounds for agents to determine appropriate responses as actions [113, 104, 31].”