CoChat
Sign Up
Cited by 4Topic ModelingMulti-Agent Systems and NegotiationNatural Language Processing Techniques

GroupDebate: Enhancing the Efficiency of Multi-Agent Debate Using Group Discussion

Open access
Liu, Tongxuan, Wang, Xingyu, Huang, Weizhe, Xu, Wenjiang, Zeng, Yuting, Jiang, Lei, Yang, Hailong, Li, Jing·2026·Proceedings of the 25th International Conference on Autonomous Agents and Multiagent Systems
DOI 10.65109/dsdx8860
Open this paper in CoChat — free accountOpen PDF

Abstract

In recent years, Large Language Models (LLMs) have demonstrated remarkable capabilities across diverse NLP tasks, including complex logical reasoning, mathematical problem-solving, and multi-step decision-making. Extensive research has explored how to enhance the logical reasoning abilities such as Chain-of-Thought, Chain-of-Thought with Self-Consistency, Tree-Of-Thoughts, and multi-agent debates. In the context of multi-agent debates, significant performance improvements can be achieved with an increasing number of agents and debate rounds. However, the escalation in the number of agents and debate rounds can drastically raise the tokens cost of debates, thereby limiting the scalability of the multi-agent debate technique. To better harness the advantages of multi-agent debates in logical reasoning tasks, this paper proposes a method to significantly reduce token cost in multi-agent debates. This approach involves dividing all agents into multiple debate groups, with agents engaging in debates within their respective groups and sharing interim debate results between groups. Comparative experiments across multiple datasets have demonstrated that this method can reduce the total tokens by up to 46.9% during debates and while potentially enhancing accuracy by as much as 21.9%. Our method significantly enhances the performance and efficiency of interactions in the multi-agent debate.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.65109/dsdx8860

Cites

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Structured contrast and deliberation among agents can enhance argumentative coherence and mitigate bias.

    “Other studies have demonstrated that structured contrast and deliberation among agents can enhance argumentative coherence and mitigate bias (Chen et al., 2024;Liu et al., 2026).”

    From Multi-Agent debate system based on large language models: structured deliberation and validation in satellite communications (2026) · cites this paper as Liu et al., 2026

    The Preliminaries describe agents reviewing and responding to one another, and the Conclusion reports improved accuracy on logical-reasoning datasets. That supports a benefit from structured deliberation, but the paper does not establish improved argumentative coherence or reduced bias specifically; accuracy and diverse viewpoints are not measures of either outcome. (LLM, conf=0.84)

    Experimental — accuracy not yet measured

  • Most existing multi-agent frameworks replicate a single LLM across all agents.

    “most existing frameworks still replicate a single LLM across all agents (Chen et al., 2025;Liu et al., 2024).”

    From Multi-Agent debate system based on large language models: structured deliberation and validation in satellite communications (2026) · cites this paper as Liu et al., 2024

    Section 2.1 describes multi-agent debate as involving multiple LLM-based agents, and the Implementation Details list models used in the paper’s experiments. Neither passage says that most existing frameworks replicate one LLM across all agents. Searches for same-model agents and replication found no passage establishing that prevalence claim; page numbers were not available in our copy. (LLM, conf=0.72)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

No references could be resolved yet.

Research works printed but not linked: 40

  1. Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Flo- rencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shya- mal Anadkat, et al. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774 (2023). · arXiv
  2. Rohan Anil, Andrew M Dai, Orhan Firat, Melvin Johnson, Dmitry Lepikhin, Alexandre Passos, Siamak Shakeri, Emanuel Taropa, Paige Bailey, Zhifeng Chen, et al. 2023. Palm 2 technical report. arXiv preprint arXiv:2305.10403 (2023). · arXiv
  3. Jonas Becker, Lars Benedikt Kaesberg, Andreas Stephan, Jan Philip Wahle, Terry Ruas, and Bela Gipp. 2025. Stay Focused: Problem Drift in Multi-Agent Debate. arXiv:2502.19559 [cs.CL] https://arxiv.org/abs/2502.19559 · arXiv
  4. Maciej Besta, Nils Blach, Ales Kubicek, Robert Gerstenberger, Michal Pod- stawski, Lukas Gianinazzi, Joanna Gajda, Tomasz Lehmann, Hubert Niewiadom- ski, Piotr Nyczyk, et al. 2024. Graph of thoughts: Solving elaborate problems with large language models. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 38. 17682-17690.
  5. Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, et al. 2020. Language models are few-shot learners. Advances in neural information processing systems 33 (2020), 1877-1901.
  6. Sébastien Bubeck, Varun Chandrasekaran, Ronen Eldan, Johannes Gehrke, Eric Horvitz, Ece Kamar, Peter Lee, Yin Tat Lee, Yuanzhi Li, Scott Lundberg, et al. 2023. Sparks of artificial general intelligence: Early experiments with gpt-4. arXiv preprint arXiv:2303.12712 (2023). · arXiv
  7. Chi-Min Chan, Weize Chen, Yusheng Su, Jianxuan Yu, Wei Xue, Shanghang Zhang, Jie Fu, and Zhiyuan Liu. 2023. Chateval: Towards better llm-based eval- uators through multi-agent debate. arXiv preprint arXiv:2308.07201 (2023). · arXiv
  8. Aakanksha Chowdhery, Sharan Narang, Jacob Devlin, Maarten Bosma, Gaurav Mishra, Adam Roberts, Paul Barham, Hyung Won Chung, Charles Sutton, Se- bastian Gehrmann, et al. 2023. Palm: Scaling language modeling with pathways. Journal of Machine Learning Research 24, 240 (2023), 1-113.
  9. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, et al. 2021. Training verifiers to solve math word problems. arXiv preprint arXiv:2110.14168 (2021). · arXiv
  10. Yilun Du, Shuang Li, Antonio Torralba, Joshua B Tenenbaum, and Igor Mordatch. 2023. Improving factuality and reasoning in language models through multia- gent debate. arXiv preprint arXiv:2305.14325 (2023). · arXiv
  11. Yao Fu, Hao Peng, Tushar Khot, and Mirella Lapata. 2023. Improving language model negotiation with self-play and in-context learning from ai feedback. arXiv preprint arXiv:2305.10142 (2023). · arXiv
  12. Taicheng Guo, Xiuying Chen, Yaqi Wang, Ruidi Chang, Shichao Pei, Nitesh V. Chawla, Olaf Wiest, and Xiangliang Zhang. 2024. Large Language Model based Multi-Agents: A Survey of Progress and Challenges. arXiv:2402.01680 [cs.CL] · arXiv
  13. Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. 2020. Measuring massive multitask language under- standing. arXiv preprint arXiv:2009.03300 (2020). · arXiv
  14. Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, Steven Basart, Eric Tang, Dawn Song, and Jacob Steinhardt. 2021. Measuring mathematical problem solving with the math dataset. arXiv preprint arXiv:2103.03874 (2021). · arXiv
  15. Lei Huang, Weijiang Yu, Weitao Ma, Weihong Zhong, Zhangyin Feng, Haotian Wang, Qianglong Chen, Weihua Peng, Xiaocheng Feng, Bing Qin, et al. 2023. A survey on hallucination in large language models: Principles, taxonomy, chal- lenges, and open questions. arXiv preprint arXiv:2311.05232 (2023). · arXiv
  16. Ziwei Ji, Nayeon Lee, Rita Frieske, Tiezheng Yu, Dan Su, Yan Xu, Etsuko Ishii, Ye Jin Bang, Andrea Madotto, and Pascale Fung. 2023. Survey of hallucination in natural language generation. Comput. Surveys 55, 12 (2023), 1-38.
  17. Ziqi Jin and Wei Lu. 2023. Tab-cot: Zero-shot tabular chain of thought. arXiv preprint arXiv:2305.17812 (2023). · arXiv
  18. Takeshi Kojima, Shixiang Shane Gu, Machel Reid, Yutaka Matsuo, and Yusuke Iwasawa. 2022. Large language models are zero-shot reasoners. Advances in neural information processing systems 35 (2022), 22199-22213.
  19. Yunxuan Li, Yibing Du, Jiageng Zhang, Le Hou, Peter Grabowski, Yeqing Li, and Eugene Ie. 2024. Improving Multi-Agent Debate with Sparse Communication Topology. arXiv preprint arXiv:2406.11776 (2024). · arXiv
  20. Tian Liang, Zhiwei He, Wenxiang Jiao, Xing Wang, Yan Wang, Rui Wang, Yujiu Yang, Zhaopeng Tu, and Shuming Shi. 2023. Encouraging divergent thinking in large language models through multi-agent debate. arXiv preprint arXiv:2305.19118 (2023). · arXiv
  21. Hunter Lightman, Vineet Kosaraju, Yura Burda, Harri Edwards, Bowen Baker, Teddy Lee, Jan Leike, John Schulman, Ilya Sutskever, and Karl Cobbe. 2023. Let's Verify Step by Step. arXiv preprint arXiv:2305.20050 (2023). · arXiv
  22. Zijie Lin and Bryan Hooi. 2025. Enhancing Multi-Agent Debate System Perfor- mance via Confidence Expression. arXiv:2509.14034 [cs.CL] https://arxiv.org/ abs/2509.14034 · arXiv
  23. Hanmeng Liu, Ruoxi Ning, Zhiyang Teng, Jian Liu, Qiji Zhou, and Yue Zhang. 2023. Evaluating the logical reasoning ability of chatgpt and gpt-4. arXiv preprint arXiv:2304.03439 (2023). · arXiv
  24. Xuefei Ning, Zinan Lin, Zixuan Zhou, Huazhong Yang, and Yu Wang. 2023. Skeleton-of-thought: Large language models can do parallel decoding. arXiv preprint arXiv:2307.15337 (2023). · arXiv
  25. Maxwell Nye, Anders Johan Andreassen, Guy Gur-Ari, Henryk Michalewski, Jacob Austin, David Bieber, David Dohan, Aitor Lewkowycz, Maarten Bosma, David Luan, et al. 2021. Show your work: Scratchpads for intermediate compu- tation with language models. arXiv preprint arXiv:2112.00114 (2021). · arXiv
  26. Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, Ilya Sutskever, et al. 2019. Language models are unsupervised multitask learners. OpenAI blog 1, 8 (2019), 9.
  27. David Rein, Betty Li Hou, Asa Cooper Stickland, Jackson Petty, Richard Yuanzhe Pang, Julien Dirani, Julian Michael, and Samuel R. Bowman. 2023. GPQA: A Graduate-Level Google-Proof Q&A Benchmark. arXiv:2311.12022 [cs.AI] https: //arxiv.org/abs/2311.12022 · arXiv
  28. KaShun Shum, Shizhe Diao, and Tong Zhang. 2023. Automatic prompt augmen- tation and selection with chain-of-thought from labeled data. arXiv preprint arXiv:2302.12822 (2023). · arXiv
  29. Qiushi Sun, Zhangyue Yin, Xiang Li, Zhiyong Wu, Xipeng Qiu, and Lingpeng Kong. 2023. Corex: Pushing the boundaries of complex reasoning through multi- model collaboration. arXiv preprint arXiv:2310.00280 (2023). · arXiv
  30. Mikhail Terekhov, Romain Graux, Eduardo Neville, Denis Rosset, and Gabin Kolly. 2023. Second-order Jailbreaks: Generative Agents Successfully Manip- ulate Through an Intermediary. In Multi-Agent Security Workshop@ NeurIPS'23.
  31. Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, et al. 2023. Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971 (2023). · arXiv
  32. Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yas- mine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhos- ale, et al. 2023. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288 (2023). · arXiv
  33. Haotian Wang, Xiyuan Du, Weijiang Yu, Qianglong Chen, Kun Zhu, Zheng Chu, Lian Yan, and Yi Guan. 2023. Apollo's Oracle: Retrieval-Augmented Reasoning in Multi-Agent Debates. arXiv preprint arXiv:2312.04854 (2023). · arXiv
  34. Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc Le, Ed Chi, Sharan Narang, Aakanksha Chowdhery, and Denny Zhou. 2022. Self-consistency improves chain of thought reasoning in language models. arXiv preprint arXiv:2203.11171 (2022). · arXiv
  35. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, et al. 2022. Chain-of-thought prompting elicits reason- ing in large language models. Advances in neural information processing systems 35 (2022), 24824-24837.
  36. Kai Xiong, Xiao Ding, Yixin Cao, Ting Liu, and Bing Qin. 2023. Examining inter- consistency of large language models collaboration: An in-depth analysis via debate. In The 2023 Conference on Empirical Methods in Natural Language Pro- cessing.
  37. Zhenran Xu, Senbao Shi, Baotian Hu, Jindi Yu, Dongfang Li, Min Zhang, and Yuxiang Wu. 2023. Towards reasoning in large language models via multi-agent peer review collaboration. arXiv preprint arXiv:2311.08152 (2023). · arXiv
  38. Shunyu Yao, Dian Yu, Jeffrey Zhao, Izhak Shafran, Tom Griffiths, Yuan Cao, and Karthik Narasimhan. 2024. Tree of thoughts: Deliberate problem solving with large language models. Advances in Neural Information Processing Systems 36 (2024).
  39. Eric Zelikman, Yuhuai Wu, Jesse Mu, and Noah Goodman. 2022. Star: Boot- strapping reasoning with reasoning. Advances in Neural Information Processing Systems 35 (2022), 15476-15488.
  40. Yongchao Zhou, Andrei Ioan Muresanu, Ziwen Han, Keiran Paster, Silviu Pitis, Harris Chan, and Jimmy Ba. 2022. Large language models are human-level prompt engineers. arXiv preprint arXiv:2211.01910 (2022). · arXiv

Cited by

Sorted by citations

Authors
Liu, T., Wang, X., Huang, W., Xu, W., Zeng, Y., Jiang, L., Yang, H., & Li, J.
Affiliations
  • University of Science and Technology of China (CN)
  • Chinese Academy of Sciences (CN)
  • Jingdong (China) (CN)
  • Beihang University (CN)
ORCID
Venue
Proceedings of the 25th International Conference on Autonomous Agents and Multiagent Systems
Publisher
International Foundation for Autonomous Agents and Multiagent Systems
Published
2026-05-24
ISSN
1548-8403 (ISSN-L)
DOI
10.65109/dsdx8860
OpenAlex
W4403753151
Cited by
4
References
0
Citations by year
2026: 3 · 2025: 1
FWCI
25.64
Type
conference paper
Language
en
Open access
Licence
cc-by
Locations
  • Adaptive Agents and Multi-Agents Systems (publishedVersion)
  • arXiv (Cornell University) (submittedVersion, cc-by)
  • arXiv (Cornell University) (cc-by)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
  • Physical Sciences › Computer Science › Artificial Intelligence › Multi-Agent Systems and Negotiation
  • Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
Keywords
multi-agent debate, group debate, token cost reduction, large language models, logical reasoning, token efficiency, group discussion, chain-of-thought

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review