📄 中文摘要
该研究提出了一种基于多智能体模型的计算数学发现新方法，旨在通过实验、证明努力和反例的相互作用来揭示数学概念。该系统自主提出猜想，并尝试进行证明，依据反馈和不断演变的数据分布做出决策。研究以欧拉多面体猜想的历史和文献中的开放挑战为灵感，基准测试任务为从多面体数据和线性代数知识中自主恢复同调概念。实验结果表明，该系统成功完成了这一学习任务，展示了其在数学发现中的潜力。
通过多智能体系统发现数学概念
出处: Discovering mathematical concepts through a multi-agent system
发布: 2026年3月6日
📄 English Summary
Discovering mathematical concepts through a multi-agent system
This research presents a novel multi-agent model for computational mathematical discovery, aimed at revealing mathematical concepts through the interplay of experimentation, proof efforts, and counterexamples. The system autonomously poses conjectures and attempts to prove them, making decisions informed by feedback and an evolving data distribution. Inspired by the historical context of Euler's conjecture for polyhedra and an open challenge in the literature, the benchmarking task focuses on autonomously recovering the concept of homology from polyhedral data and knowledge of linear algebra. Experimental results demonstrate that the system successfully completes this learning problem, showcasing its potential in mathematical discovery.