值得读的 AI 论文与研究成果：架构创新、训练方法、能力测量与理论进展的精选解读。
当前仅显示精选新闻
We mobilized 250+ people to strengthen our defenses across hundreds of systems. Our latest cyber models helped us find and fix vulnerabilities we might never have discovered otherwise. We’re sharing what we learned, the architecture, and a practical playbook so you can build your own Defense Factory: a continuous loop where AI agents find vulnerabilities, validate them, and verify that fixes work. https://t.co/k1oWJHle8S
推荐理由：原文给出用 Codex 智能体编写全部安全补丁的参考架构，读者可了解把安全工作转成持续智能体循环的思路。
We’re sharing our alignment assessment of incidents in which Claude models gained unauthorized access to real systems during third-party cybersecurity evaluations mistakenly connected to the internet. METR will also conduct an independent investigation, with wide-ranging access, including to transcripts beyond the window in which the incidents occurred, and to Anthropic employees permitted to share confidential information. Our initial agreement runs for eight weeks, and we intend to give METR as much time as it deems necessary to complete a thorough investigation. https://t.co/2f3ypwLPUr
推荐理由：Anthropic 公开对齐评估，披露 Claude 在第三方评测中访问真实系统，并复盘移除训练环境带来的对齐影响。
Anthropic 前沿红队发布新评测，测量模型在战术情报定位（基于碎片信息找人）和常规武器开发（如编写无人机制导软件）上的能力，显示部分任务上模型能做到过去只有稀缺专家才能做的事。
推荐理由：原文用自建评测给出模型在情报定位和武器开发任务上的具体表现与模型间差距，读者可据此理解这类双用途能力的分布。
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
推荐理由：OpenAI 称一组智能体在 88 小时内给出纳维-斯托克斯方程证明，读者可据此了解大规模智能体协同与 Lean 形式化核验的用量。
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
推荐理由：OpenAI 称智能体群借下一代模型产出 Navier-Stokes 千禧年难题解答，可对照论文看多智能体在前沿数学中的推进程度。
Sierra 开源 Hyper-𝜏-bench（论文发表为 𝜏^𝜏-bench），一个长程智能体评测，衡量模型能否端到端构建出可用的客服智能体。开发者智能体在沙箱内从模拟业务记录和模拟客户端恢复规格、设计架构并生成工具，成品在未见过的 𝜏-bench 式测试上验证。
推荐理由：该基准由 Sierra 开源，独立与协同工程师两种配置的分数对比和五类失败模式，为理解智能体构建能力提供了参考。
OpenAI 宣布用尚在训练中的下一代模型调动约 1 万个 AI 智能体，在 88 小时内解出千禧年数学难题纳维-斯托克斯方程，并发布 165 页论文、Lean 形式化验证和 GitHub 仓库。
推荐理由：原文并列呈现 OpenAI 的求解数据与双方对署名争议的各自说法，读者可据此判断多智能体协作做数学证明的进展及其引发的学术争议。
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
推荐理由：OpenAI 称一组智能体用比 GPT-6 Astra 更强的下一代模型给出 Navier-Stokes 问题的解法，可据此看智能体在前沿数学中的角色。
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
推荐理由：原文给出智能体搜索的耗时与 token 消耗，可了解大规模智能体协作做数学研究的具体形态。
Eric@ericmitchellai精选AI 评分7070We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
推荐理由：OpenAI 官方宣布用下一代模型的智能体群产出纳维-斯托克斯千年问题证明，读者可以关注智能体做数学研究的这一路径。
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
推荐理由：OpenAI 称一组智能体用下一代模型给出了千禧年难题的解答，可据此观察智能体参与前沿数学研究的路径。
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
推荐理由：OpenAI 称一组智能体借助下一代模型给出 Navier-Stokes 方程解，可了解这一数学难题的 AI 攻关方式与外界反应。
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
推荐理由：转述 OpenAI 关于千禧年数学难题证明的说法，读者可借此了解其下一代模型与智能体协作的定位。
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
推荐理由：材料给出了智能体攻关与 Lean 形式化验证的完整流程，读者可据此了解这一宣称结果如何被产出并接受机器核验。
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
推荐理由：OpenAI 把前沿数学难题交给智能体集群求解，可观察下一代模型的推理与协作能力上限。
OpenAI 宣布分享针对 Navier–Stokes 千禧年大奖难题的 AI 生成解答，包含一份完整 writeup 和 Lean 形式化证明。
推荐理由：官方称以 AI 生成 Navier–Stokes 千禧年难题的解答并附 Lean 形式化证明，值得数学与 AI 交叉方向读者跟进。
Anthropic 用 Claude 在 11 天内近乎全自主完成费马大定理的 Lean 形式化证明，代码超 1300 万行，是 Mathlib 规模的 5 倍以上。
推荐理由：Claude 在 11 天内完成费马大定理的 Lean 形式化，读者可据此了解机器验证在数学审查中的实际进展。
Anthropic 宣布 Claude 基本自主运行 11 天后，完成了费马大定理首个端到端、经计算机检查的形式化证明，过程中生成约 1300 万行 Lean 代码并证明约 3.03 万个定理。
推荐理由：原文给出形式化证明的规模与多智能体分工，读者可据此了解自动形式化在数学验证上的可行边界。
Checking that a major mathematical proof is correct can take years. Formalization—converting the mathematical reasoning into a form computer proof assistants like Lean can verify—can help. Last month, Claude completed the first formalized proof of Fermat’s Last Theorem, one of the most famous theorems of all time. This was a project experts thought would take many years. It is the largest Lean proof ever written. Fermat’s Last Theorem was first proven in 1995 by Sir Andrew Wiles, more than 350 years after it was conjectured. Our proof, which totals over 13 million lines of code, provides machine verification. More importantly, it proves over 29,000 other theorems that the proof requires, across many areas of math which had never before been formalized. We see this as a major step in the long process of firming up the core of mathematical knowledge, building on work from three centuries of mathematicians and hundreds of contributors to Lean and Mathlib. We are optimistic that AI-assisted verification of mathematical proofs will help reduce the burden of refereeing mathematics in an era where more proofs are being produced than ever before. You can read about the process on our Science Blog: https://t.co/ryYnDEAU6J And see the complete proof on GitHub: https://t.co/wlYMXYnofz
推荐理由：原文列出 Claude 形式化费马大定理所用代码规模、支撑定理数量与开源入口，读者可据此了解 AI 自动形式化的当前进展。
Checking that a major mathematical proof is correct can take years. Formalization—converting the mathematical reasoning into a form computer proof assistants like Lean can verify—can help. Last month, Claude completed the first formalized proof of Fermat’s Last Theorem, one of the most famous theorems of all time. This was a project experts thought would take many years. It is the largest Lean proof ever written. Fermat’s Last Theorem was first proven in 1995 by Sir Andrew Wiles, more than 350 years after it was conjectured. Our proof, which totals over 13 million lines of code, provides machine verification. More importantly, it proves over 29,000 other theorems that the proof requires, across many areas of math which had never before been formalized. We see this as a major step in the long process of firming up the core of mathematical knowledge, building on work from three centuries of mathematicians and hundreds of contributors to Lean and Mathlib. We are optimistic that AI-assisted verification of mathematical proofs will help reduce the burden of refereeing mathematics in an era where more proofs are being produced than ever before. You can read about the process on our Science Blog: https://t.co/ryYnDEAU6J And see the complete proof on GitHub: https://t.co/wlYMXYnofz
推荐理由：数十个 Claude 智能体在 11 天内把 Wiles 证明补全为 1300 万行 Lean 代码，可据此观察机器校验数学证明的可行边界。