Rohan Paul· @rohanpaul_ai · X·· 3 小时前精选AI 评分73
AI 导读
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 台此前没有文献分析过的 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 证明检查器中完成规则证明，这些规则均通过 Lean 验证。
推荐理由
这篇论文展示 Gemini 智能体从回答人类提出的数学问题转向开放式发现，并给出 Lean 验证的具体路径。
正文 · 原文
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