Google DeepMind 与剑桥的新论文显示，一个基于 Gemini 的智能体能够自主发现新数学，为 11 台此前无分析记录的 Busy Beaver 图灵机找到并给出精确规则。该智能体在模拟器中测试边界情况、起草非形式化证明，再在 Lean 证明检查器中完成验证。论文题为 At the Limits of Computability: Steps Towards Autonomous AI Discovery on the Busy Beaver Frontier，作者来自 Google DeepMind 与剑桥大学。
论文展示 Gemini 智能体在 Busy Beaver 问题上自主发现并形式化验证新规则，可作为开放式数学发现的案例参照。
New Google DeepMind and Cambridge paper shows that an AI agent can discover new math itself, proving exact rules for 11 unsolved Turing machines.
AI can move from answering human-posed math questions to open-ended discovery, shown by a Gemini-based agent that found and Lean-verified new rules for 11 Busy Beaver machines with no prior documented analysis.
It tested edge cases in a simulator, drafts informal proofs, and then proves the rules in the Lean proof checker.
来源：Rohan Paul · x.com