arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2311.08152cs.CL
收藏

通过多智能体同行评审协作实现大语言模型的推理能力

Towards Reasoning in Large Language Models via Multi-Agent Peer Review Collaboration

Harbin Institute of Technology (Shenzhen)(哈尔滨工业大学(深圳));University College London(伦敦大学学院)

Zhenran Xu, Senbao Shi, Baotian Hu, Jindi Yu, Dongfang Li, Min Zhang, Yuxiang Wu

AI总结:

针对大语言模型复杂推理能力不足的问题,提出模拟学术同行评审的多智能体协作策略,通过独立解题、带置信度的互评修正提升推理效果,在多任务多数据集上准确率优于现有方法。

摘要:

大语言模型(LLMs)在通用自然语言处理任务中展现出了卓越的能力,但在复杂推理任务中往往表现不足。近期研究探索了类人的问题解决策略,例如自我纠错,以进一步拓展单模型推理能力的边界。在本研究中,我们通过让多个模型互相纠错,使单个模型能够“跳出思维定式”。我们提出了一种模拟学术同行评审流程的多智能体协作策略。每个智能体独立构建自己的解决方案,对其他智能体的解决方案给出评审意见,并为自己的评审赋予置信度水平。在收到同行评审后,各智能体修正其初始解决方案。在三类不同推理任务上开展的大量实验表明,与现有方法相比,我们的协作方法在全部十个数据集上都实现了更优的准确率。进一步研究证实了在评审中融入置信度的有效性,证明了反馈交换优于单纯的方案共享,并强调了能力与多样性对促成成功协作的作用。

英文摘要:

Large Language Models (LLMs) have shown remarkable capabilities in general natural language processing tasks but often fall short in complex reasoning tasks. Recent studies have explored human-like problem-solving strategies, such as self-correct, to push further the boundary of single-model reasoning ability. In this work, we let a single model "step outside the box" by engaging multiple models to correct each other. We introduce a multi-agent collaboration strategy that emulates the academic peer review process. Each agent independently constructs its own solution, provides reviews on the solutions of others, and assigns confidence levels to its reviews. Upon receiving peer reviews, agents revise their initial solutions. Extensive experiments on three different types of reasoning tasks show that our collaboration approach delivers superior accuracy across all ten datasets compared to existing methods. Further study underscores the effectiveness of integrating confidence in reviews, demonstrates the superiority of feedback exchange over mere solution sharing, and highlights the role of capability and diversity in fostering successful collaboration.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑