用多智能体协作让AI更准地理解表格，避免算错和瞎编。
Mixture-of-Minds: Multi-Agent Reinforcement Learning for Table Understanding
- 分三个角色：规划、写代码、回答，各司其职。
- 用强化学习训练，自动生成高质量推理路径，准确率达62.13%。
- 适合需要精准表格推理的金融、科研场景。
表格理解和推理在诸多现实应用中至关重要。大语言模型（LLMs）虽有潜力，但现有方法仍有局限：微调类方法强化语言推理却易出算术错误和幻觉；工具类方法能精确操作表格，但依赖僵化模式且缺乏语义理解。为此，本文提出Mixture-of-Minds，一种多智能体框架，将表格推理分解为规划、编码、回答三个专门角色，使每个智能体专注特定任务，并通过代码执行实现精确表操作。基于此流程，我们设计自提升训练框架，利用蒙特卡洛树搜索（MCTS）生成伪黄金轨迹，以强化学习优化各智能体。大量实验表明，Mixture-of-Minds在TableBench上达到62.13%准确率，超越OpenAI-o4-mini-high。结果证明，结合结构化多智能体流程与强化学习，可有效推动表格理解发展。
原文摘要 · Abstract (English)
Understanding and reasoning over tables is a critical capability for many real-world applications. Large language models (LLMs) have shown promise on this task, but current approaches remain limited. Fine-tuning based methods strengthen language reasoning; yet they are prone to arithmetic errors and hallucination. In contrast, tool-based methods enable precise table manipulation but rely on rigid schemas and lack semantic understanding. These complementary drawbacks highlight the need for approaches that integrate robust reasoning with reliable table processing. In this work, we propose Mixture-of-Minds, a multi-agent framework that decomposes table reasoning into three specialized roles: planning, coding, and answering. This design enables each agent to focus on a specific aspect of the task while leveraging code execution for precise table manipulation. Building on this workflow, we introduce a self-improvement training framework that employs Monte Carlo Tree Search (MCTS) rollouts to generate pseudo-gold trajectories and optimize agents with reinforcement learning (RL). Extensive experiments show that Mixture-of-Minds delivers substantial gains, reaching 62.13% on TableBench and surpassing OpenAI-o4-mini-high. These results demonstrate the promise of combining structured multi-agent workflows with RL to advance table understanding.
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。