我们宣布在数学、计算机科学和物理学领域关于调和分析、复杂性理论和量子场论的三项新发现。
我们的发布流程在相关学术界的反馈下持续演进。在我们看来,这些手稿最好被理解为一种原始研究数据。为了将这些原始研究数据转化为科学理解,在本次发布中,我们正在尝试一种新方法。我们联系了外部专家,请他们担任引导者,组织团队来理解这些新结果。他们建议了三种不同的发布策略:
对于数学结果,发布AI输出,让研究人员自由组织自己的努力来解释该结果。
对于计算机科学结果,组建一个外部合作团队来解释该结果,同时将AI输出公开,供任何人检查和在此基础上继续工作。
对于物理学结果,组建一个外部合作团队,并在材料发布上听从他们的意见。
我们尚不知道这三种方法中哪一种最有用,我们欢迎就哪些方法有效以及我们可以改进哪些方面提供反馈。
外部合作将遵循四项原则:
透明度。 我们对我们合作的存在保持透明。参与者不会被要求签署保密协议,并将可以自由讨论工作、分享中间成果并发表他们的结论。
不寻求学术荣誉。 我们不为我们内部模型产生的发现寻求学术荣誉。邀请研究人员帮助管理合作也不意味着他们或他们的工作组做出了原始发现。
认可。 该团队应因其形成的理解、撰写的文稿以及创造的其他材料而获得认可。该团队还将帮助确保工作适当归功于其所借鉴的研究人员、技术和想法。
独立性。 参与是自愿且无报酬的。工作组成员独立参与,他们的结论是他们自己的。他们不代表OpenAI发言,OpenAI也不代表他们发言。
我们将把与这些结果相关的材料和产物移交给合作团队。对于计算机科学合作,该团队已决定立即公开自然语言证明和Lean形式化,我们将直接将其发布到我们的GitHub仓库。对于物理学合作,该团队已决定暂不发布AI生成的产物,而是自行撰写。
所有这些方法都涉及权衡。一方面,合作使专家能够促进对结果的解读,并将其置于更广泛的学术背景中。另一方面,选择初始引导者必然会影响谁首先参与,并造成信息不对称。我们欢迎就如何扩大参与并使这一过程更有用提供反馈。
局部平滑与Kakeya猜想
数学领域的结果涉及调和分析。1991年,Sogge提出了局部平滑猜想,该猜想涉及波在聚焦时的行为。特别是,该猜想指出,最极端的聚焦必须是短暂的。在典型随机时间上的聚焦量总是远小于在最坏情况时间上可能发生的情况。这是整个数学中一个熟悉的模式:最坏情况下可能发生的情况比平均后观察到的情况更极端。由于这一猜想现象的根本性,它也在整个数学分析中具有广泛的影响。例如,它蕴含Kakeya猜想和Stein的限制猜想。
2019年,Guth、Wang和Zhang(在新窗口中打开)证明了二维空间中的局部平滑猜想。2025年,Wang和Zahl(在新窗口中打开)证明了三维中的Kakeya猜想。我们的模型给出了局部平滑猜想在所有维度上的证明,因此也解决了所有维度中的Kakeya猜想和Stein的限制猜想。新证明基于一个新的辅助偏微分方程。
我们在2026年10月18日获得了AI输出,并于次日用Lean完成了形式化,随后我们联系了一位专家,讨论围绕这一结果组建合作。根据他们的建议,我们决定公开AI输出,对这一结果采用与我们2026年10月6日发布相同的策略。我们已在公共GitHub仓库中发布了由我们的模型生成的AI自然语言输出文稿(在新窗口中打开)和Lean证明(在新窗口中打开)。
指数时间假说为假
计算机科学合作将理解并解释一个关于指数时间假说(ETH)的结果,这是一个关于布尔公式可满足性计算复杂性的长期猜想,该问题与P vs. NP问题密切相关。ETH一直被广泛认为是真的,并已在数百篇论文中用于推导图论、计算生物学、数据库、计算几何等广泛领域的条件计算困难性结果。令人惊讶的是,我们的系统最近证明了ETH为假,这一结果随后在Lean中得到了验证。对ETH的反驳重新开启了关于这些任务能变得多快的问题,并可能有助于指导寻找更好的算法。找到这样的算法,并将理论进展转化为实际改进,将需要大量进一步研究。
这项工作将由Russell Impagliazzo(加州大学圣地亚哥分校)、Ramamohan Paturi(加州大学圣地亚哥分校)和Ryan Williams(麻省理工学院)领导的团队进行验证和引导。这三位都是算法设计和复杂性理论方面的专家。
我们已同意该团队的要求,即所有结果的原始数据都应公开,以便社区进行独立验证,并且OpenAI承诺资助一次会议,以理解这项工作对整个领域的影响。该团队邀请社区帮助组织对ETH主张的公开审查,并讨论AI生成的突破在验证、获取、荣誉和研究未来方面所引发的更广泛问题。
我们的AI于2026年10月16日为其ETH找到了自然语言反证,随后我们于2026年10月17日在Lean中形式化了该结果,这促使我们开展了此处描述的合作。为此,我们在公共GitHub仓库中发布了由我们的模型生成的AI自然语言输出报告(在新窗口中打开)和一份Lean证明(在新窗口中打开)。
计算复杂性领域有许多开放问题和猜想,迄今为止,科学家们竭尽全力仍未能证明或反驳它们。我们希望人类与AI能够为该领域提供新的理解,并揭示高效计算的内在能力与局限。
量子色动力学的存在性与标准模型的完备化
物理学合作将理解并解释一项结果,该结果为四维中包含规范场和费米子的某些相互作用量子场论提供了数学严格的存在性证明,包括量子色动力学(QCD)以及粒子物理标准模型的统一完备化。我们的工作建立在Balaban的方法(在新窗口中打开)之上,属于构造性量子场论。作为一个子问题,该工作还包括纯Yang–Mills理论存在性的证明,这是量子Yang–Mills理论克雷数学研究所千禧年大奖难题(在新窗口中打开)的一部分。
标准模型描述了四种基本力中三种(不包括引力)的基本亚原子粒子相互作用。该存在性证明涉及最简单的大统一理论,该理论包含标准模型的规范场和费米子(因为标准模型本身无法外推到任意短的距离)。QCD描述了夸克和胶子等粒子通过强相互作用的相互作用。纯Yang–Mills理论仅描述胶子(力载体)的自相互作用,而不涉及夸克等物质。
David Gross将领导一个包括Michael Douglas、Arthur Jaffe、Igor Klebanov和Nikita Nekrasov在内的小组来指导这项工作。
Gross于1973年在Yang-Mills理论中关于渐近自由的工作帮助确立了QCD作为强相互作用的理论,并为他赢得了2004年诺贝尔物理学奖(在新窗口中打开)的份额。渐近自由也是这些理论作为数学严格对象存在性的关键机制。
Jaffe与Witten共同提出了量子Yang-Mills理论千禧年大奖难题(在新窗口中打开),并且还构想了千禧年大奖难题的想法,并共同创立了克雷数学研究所。
Douglas在过去20年中一直是该问题的管理者,于2004年报告了其状态(在新窗口中打开),然后于今年年初提供了关于进展的全面更新(在新窗口中打开)。
Klebanov帮助开创了规范-弦对偶在禁闭(QCD的一个重要性质)中的应用,现在领导Simons禁闭与QCD弦合作组织(在新窗口中打开),
Nekrasov在Yang–Mills理论推广的精确方法方面做出了开创性工作,并于2023年获得了Dannie Heineman数学物理奖。
该小组要求,与该工作相关的所有通信,无论与小组任何成员,均应发送至以下电子邮件账户:YANGMILLS4D@gmail.com。
我们的模型于2026年9月9日生成了Yang–Mills存在性的自然语言证明,随后于2026年9月18日生成了QCD的证明。鉴于工作的长度和复杂性,我们将发布推迟到Lean形式化完成之后。我们于2026年10月16日完成了Yang–Mills存在性的形式化,这促使我们开展了此处描述的合作。
我们希望这些结果标志着量子场论——现代理论物理学中最重要的框架——开始转变为一个数学严格的领域。我们很高兴为学术界主导的进展提供资源,以实现这一目标。
支持科学家
我们致力于支持理解和基于这些结果开展更广泛的工作。这些合作补充了我们最近宣布的向Simons基金会提供的500万美元捐款(在新窗口中打开)。该资金由Simons基金会管理和授予,将支持研讨会、会议和特别项目,以帮助研究人员理解并基于AI产生的结果开展工作。我们致力于确保为研究人员提供资金,以独立审查、解释和基于上述结果开展工作。
我们继续邀请各方就披露结果和与科学家互动以加速发现的最佳方式提供反馈。
We are announcing three new discoveries in mathematics, computer science, and physics about harmonic analysis, complexity theory, and quantum field theory.
Our release process continues to evolve in response to feedback from the associated academic communities. In our view, these manuscripts are best understood as a form of raw research data. To convert this raw research data into scientific understanding, in this release we are experimenting with a new approach. We reached out to external experts and asked them to act as shepherds, organizing groups to understand the new results. They advised three different release strategies:
For the math result, release the AI output, leaving researchers free to organize their own efforts to explain the result.
For the computer science result, form an external group of collaborators to explain the result, while making the AI output publicly available for anyone to examine and build on.
For the physics result, form an external group of collaborators and defer to them on the release of materials.
We do not yet know which of these three approaches will be most useful, and we welcome feedback on what works and what we can improve.
The external collaborations will follow four principles:
Transparency. We are transparent about the existence of our collaborations. Participants will not be asked to sign nondisclosure agreements and will be free to discuss the work, share intermediate products, and publish their conclusions.
No Academic Credit. We do not seek academic credit for the discoveries produced by our internal models. Inviting a researcher to help steward a collaboration also does not imply that they or their working group made the original discovery.
**Recognition.**This group should receive recognition for the understanding they develop, the writeups they author, and the other materials they create. The group will also help ensure that the work appropriately credits researchers, techniques, and ideas that it draws on.
**Independence.**Participation is voluntary and unpaid. Working-group members participate independently, and their conclusions are their own. They do not speak on behalf of OpenAI, and OpenAI does not speak on their behalf.
We will turn over materials and artifacts related to these results to the collaborations. For the computer science collaboration, the group has decided to make the natural language proof and Lean formalization public immediately, and we are publishing them directly to our GitHub repository. For the physics collaboration, the group has decided to hold off on releasing the AI-produced artifacts and instead work on writing it up themselves.
All of these approaches involve tradeoffs. On the one hand, the collaborations empower experts to foster unpacking the results and contextualizing them in the broader academic community. On the other hand, choosing initial shepherds necessarily shapes who gets involved first and creates asymmetry of information. We welcome feedback on how to broaden participation and make this process more useful.
Local smoothing and the Kakeya conjecture
The result in mathematics concerns harmonic analysis. In 1991, Sogge proposed the local smoothing conjecture, which concerns the behavior of waves as they become focused. In particular, the conjecture states that the most extreme focusing must be short-lived. The amount of focusing at a typical random time is always much smaller than what can happen at the worst-case time. This is a familiar pattern throughout mathematics: what can happen in the worst case is more extreme than what is observed after averaging. Because of how fundamental this conjectured phenomenon is, it also has wide ranging implications throughout mathematical analysis. For instance, it implies the Kakeya conjecture and Stein’s restriction conjecture.
In 2019, Guth, Wang, and Zhang(opens in a new window) proved the local smoothing conjecture in two spatial dimensions. In 2025, Wang and Zahl(opens in a new window) proved the Kakeya conjecture in three dimensions. Our models produced a proof of the local smoothing conjecture in every dimension and therefore also resolve the Kakeya conjecture and Stein’s restriction conjecture in all dimensions. The new proof is based on a novel auxiliary partial differential equation.
We obtained our AI output on October 18th, 2026, which was formalized in Lean the following day, and then we reached out to an expert to discuss forming a collaboration around this result. Following their recommendation, we decided to release the AI output publicly, following the same strategy for this result as our October 6th, 2026 release. We have released both an AI natural language output writeup(opens in a new window) produced by our models and aLean proof (opens in a new window)in our public GitHub repository.
The Exponential Time Hypothesis is false
The computer science collaboration will understand and explain a result concerning the Exponential Time Hypothesis (ETH), a longstanding conjecture about the computational complexity of Boolean formula satisfiability, a problem deeply related to the P vs. NP problem. The ETH has been widely believed to be true, and has been used in hundreds of papers to derive conditional computational hardness results across a wide variety of fields including graph theory, computational biology, databases, computational geometry, and more. Surprisingly, our system has recently proved that the ETH is false, a result that was subsequently verified in Lean. The refutation of the ETH reopens questions about how much faster these tasks could become, and may help guide the search for better algorithms. Finding such algorithms, and turning theoretical advances into practical improvements, will require significant further research.
This work will be verified and shepherded by a group led by Russell Impagliazzo (UC San Diego), Ramamohan Paturi (UC San Diego), and Ryan Williams (MIT). All three are experts in algorithm design and complexity theory.
We have agreed to the group’s requests that all raw data of the results be made public to facilitate independent verification by the community, and that OpenAI commit to funding a conference to understand the implications of this work on the field as a whole. The group invites the community to help organize an open examination of the ETH claim and to discuss the broader questions that AI-generated breakthroughs raise about verification, access, credit, and the future of research.
Our AI found its natural language disproof for the ETH on October 16th, 2026 and then subsequently we formalized the result in Lean on October 17th, 2026, which led us to develop the collaboration described here. To that end, we have released both an AI natural language output writeup(opens in a new window) produced by our models and a Lean proof (opens in a new window)in our public GitHub repository.
Computational complexity has many open questions and conjectures, whose proofs or refutations have so far eluded the best efforts of scientists. We hope that humans with AI can provide new understanding in this field, and shed light on the inherent capabilities and limits of efficient computation.
Existence of quantum chromodynamics and completions of the Standard Model
The physics collaboration will understand and explain a result that gives a mathematically rigorous existence proof of certain interacting quantum field theories containing gauge fields and fermions in four dimensions, including Quantum Chromodynamics (QCD) and a unified completion of the Standard Model of particle physics. Our work builds on Balaban’s approach(opens in a new window) in constructive quantum field theory. As a subproblem, this work also includes a proof of the existence of pure Yang–Mills theory, which is part of the Quantum Yang–Mills Theory Clay Mathematics Institute Millennium Prize problem(opens in a new window).
The Standard Model describes the interaction of fundamental subatomic particles for three of the four fundamental forces, excepting gravity. The existence proof pertains to the simplest grand unified theory that includes the gauge fields and fermions of the Standard Model (as the Standard Model itself cannot be extrapolated to arbitrarily short distances). QCD describes the interaction of particles such as quarks and gluons via the strong force. Pure Yang–Mills theory only describes the self-interactions of the gluons, the force carriers, rather than matter such as quarks.
David Gross will lead a group including Michael Douglas, Arthur Jaffe, Igor Klebanov, and Nikita Nekrasov that will shepherd this work.
Gross’s 1973 work on asymptotic freedom in Yang-Mills theories helped establish QCD as the theory of the strong interactions and earned him a share of the 2004 Nobel Prize in Physics(opens in a new window). Asymptotic freedom is also the key mechanism underlying the existence of these theories as mathematically rigorous objects.
Jaffe formulated the Quantum Yang-Mills Theory Millenium Prize problem(opens in a new window) with Witten and also envisioned the idea of Millennium Prize problems and co-founded the Clay Mathematics Institute.
Douglas has already been a steward of this problem over the last 20 years, reporting on its status in 2004(opens in a new window) and then providing a comprehensive update on the progress(opens in a new window) at the beginning of this year.
Klebanov helped pioneer the application of gauge–string duality to confinement, an important property of QCD, and now directs the Simons Collaboration on Confinement and QCD Strings(opens in a new window),
Nekrasov has done groundbreaking work on exact methods for generalizations of Yang–Mills theory and won the Dannie Heineman Prize for Mathematical Physics in 2023.
The group requests that all communications with any and all members of the group regarding this effort be directed to the following email account: YANGMILLS4D@gmail.com.
Our models produced a natural language proof for Yang–Mills existence on September 9th, 2026, and then subsequently produced a proof for QCD on September 18th, 2026. Given the length and complexity of the work, we postponed release until the Lean formalization completed. We finished our formalization of Yang–Mills’ existence on October 16th, 2026, which led us to develop the collaboration described here.
We hope that these results mark the beginning of a transition in which quantum field theory, the most important framework in modern theoretical physics, becomes a mathematically rigorous field. We are excited to provide resources for academic-led progress towards that goal.
Supporting scientists
We are committed to supporting the broader work of understanding and building on these results. These collaborations complement our recently announced $5 million contribution(opens in a new window) to the Simons Foundation. Managed and awarded by the Simons Foundation, that funding will support workshops, conferences, and special programs to help researchers understand and build on results produced by AI. We are committed to ensuring funding for researchers to independently examine, explain, and build on the results above.
We continue to invite feedback on the best way to disclose results and engage with scientists to accelerate discovery.
本文内容采集自官方网站,排版和翻译可能与原页面存在差异。
阅读官方全文