用验证代理提升大模型推理可信度，解决多代理思维树的漏洞问题。
Improving LLM Reasoning with Multi-Agent Tree-of-Thought Validator Agent
- 多代理并行探索思维路径，验证代理筛选有效推理。
- 在GSM8K上平均比标准ToT高出5.6%准确率。
- 适合需要严谨逻辑推理的应用场景，如数学题求解。
多代理策略通过为问题求解过程分配专门角色，已成为提升大语言模型（LLMs）推理能力的有前景方法。同时，思维树（ToT）方法在复杂问答任务中展现出通过探索多样化推理路径改善推理效果的潜力。多代理推理的一个关键局限在于‘推理者’代理对推理路径的浅层探索。虽然ToT策略可缓解此问题，但可能生成错误的推理分支，损害最终答案的可信度。为结合多代理推理与ToT策略的优势，我们提出一种新方法：将基于ToT的推理者代理与思想验证代理相结合。多个推理者代理并行运行，采用ToT探索多样化推理路径；思想验证代理则对这些路径进行审查，仅保留有效推理的结论。该机制通过剔除错误推理路径，实现更稳健的投票策略，显著增强系统处理需系统性与可信推理任务的能力。在GSM8K数据集上的评估显示，该方法在四种不同LLM上平均优于标准ToT策略5.6%，表现出更优性能。代码及相关内容见：https://github.com/SecureAIAutonomyLab/MA-ToT
原文摘要 · Abstract (English)
Multi-agent strategies have emerged as a promising approach to enhance the reasoning abilities of Large Language Models (LLMs) by assigning specialized roles in the problem-solving process. Concurrently, Tree of Thoughts (ToT) methods have shown potential in improving reasoning for complex question-answering tasks by exploring diverse reasoning paths. A critical limitation in multi-agent reasoning is the 'Reasoner' agent's shallow exploration of reasoning paths. While ToT strategies could help mitigate this problem, they may generate flawed reasoning branches, which could harm the trustworthiness of the final answer. To leverage the strengths of both multi-agent reasoning and ToT strategies, we introduce a novel approach combining ToT-based Reasoner agents with a Thought Validator agent. Multiple Reasoner agents operate in parallel, employing ToT to explore diverse reasoning paths. The Thought Validator then scrutinizes these paths, considering a Reasoner's conclusion only if its reasoning is valid. This method enables a more robust voting strategy by discarding faulty reasoning paths, enhancing the system's ability to tackle tasks requiring systematic and trustworthy reasoning. Our method demonstrates superior performance compared to existing techniques when evaluated on the GSM8K dataset, outperforming the standard ToT strategy by an average 5.6% across four LLMs. The code and related content can be found in: https://github.com/SecureAIAutonomyLab/MA-ToT
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。