EUVIMEDEUVIMED entdecken
Crossref

Learning to break: Knowledge-enhanced reasoning in multi-agent debate system

Haotian Wang, Xiyuan Du, Weijiang Yu, Qianglong Chen, Kun Zhu, Zheng Chu, Lian Yan, Yi Guan

Neurocomputing · 2025-02 · 618 · 129063

DOI: 10.1016/j.neucom.2024.129063 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Neurocomputing, 2025-02. Verfasst von Haotian Wang und 7 weiteren Autor:innen. Die Arbeit wurde bisher 21-mal zitiert (Crossref) und nennt 61 Literaturangaben. Gefördert von: National Key Research and Development Program of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Multi-agent debate system (MAD) imitating the process of human discussion in pursuit of truth, aims to align the correct cognition of different agents for the optimal solution. It is challenging to make various agents perform right and highly consistent cognition due to their limited and different knowledge backgrounds (i.e., cognitive islands), which hinders the search for the optimal solution. To address the challenge, we propose a novel \underline{M}ulti-\underline{A}gent \underline{D}ebate with \underline{K}nowledge-\underline{E}nhanced framework (\textbf{MADKE}) to promote the system to find the solution. First, we involve a shared retrieval knowledge pool in the debate process to solve the problem of limited and different knowledge backgrounds. Then, we propose an adaptive knowledge selection method to guarantee the accuracy and personalization of knowledge. This method allows agents to choose whether to use external knowledge in each conversation round according to their own needs. Our experimental results on six datasets show that our method achieves state-of-the-art results compared to existing single-agent and multi-agent methods. Further analysis reveals that the introduction of retrieval knowledge can help the agent to break cognitive islands in the debate process and effectively improve the consistency and correctness of the model. Moreover, MADKE using Qwen1.5-72B-Chat surpasses GPT-4 by +1.26\% on average in six datasets, which validates that our method can help open-source LLMs achieve or even surpass the performance of GPT-4. Our code is available at \url{https://github.com/FutureForMe/MADKE}.

Förderung

  • National Key Research and Development Program of China

Zitieren

APA

Wang, H., Du, X., Yu, W., Chen, Q., Zhu, K., Chu, Z., Yan, L., & Guan, Y. (2025). Learning to break: Knowledge-enhanced reasoning in multi-agent debate system. Neurocomputing, 618, 129063. https://doi.org/10.1016/j.neucom.2024.129063

Vancouver

Wang H, Du X, Yu W, Chen Q, Zhu K, Chu Z, et al. Learning to break: Knowledge-enhanced reasoning in multi-agent debate system. Neurocomputing. 2025;618:129063. doi:10.1016/j.neucom.2024.129063

BibTeX
@article{Wang2025,
  author = {Wang, Haotian and Du, Xiyuan and Yu, Weijiang and Chen, Qianglong and Zhu, Kun and Chu, Zheng and Yan, Lian and Guan, Yi},
  title = {Learning to break: Knowledge-enhanced reasoning in multi-agent debate system},
  journal = {Neurocomputing},
  year = {2025},
  volume = {618},
  pages = {129063},
  doi = {10.1016/j.neucom.2024.129063}
}
Kostenlos starten →