







Palomar – a registry of Lean verified mathematics
In recent months there has been a proliferation of AI-generated proofs of various old and new results, some of which have been formalized in the proof assistant language Lean. However, checking tha…
Mathematics with large language models as provers and verifiers
During 2024 and 2025 the discussion about the theorem-proving capabilities of large language models started reporting interesting success stories, mostly to do with difficult exercises (such as problems from the International Mathematical Olympiad), but also with conjectures [Feldman & Karbasi, arXiv:2509.18383v1] formulated for the purpose of verifying whether the artificial intelligence could prove it. In this paper we report a theorem proving feat achieved by ChatGPT by using a protocol involving different prover and verifier instances of the gpt-5 model working collaboratively. To make sure that the produced proofs do not suffer from hallucinations, the final proof is formally verified by the lean proof assistant, and the conformance of premises and conclusion of the lean code is verified by a human. Our methodology is by no means complete or exact. It was nonetheless able to solve five out of six 2025 IMO problems, and close about a third of the sixty-six number theory conjectures in [Cohen, Journal of Integer Sequences, 2025].

Mathematics with large language models as provers and verifiers
During 2024 and 2025 the discussion about the theorem-proving capabilities of large language models started reporting interesting success stories, mostly to do with difficult exercises (such as problems from the International Mathematical Olympiad), but also with conjectures [Feldman & Karbasi, arXiv:2509.18383v1] formulated for the purpose of verifying whether the artificial intelligence could prove it. In this paper we report a theorem proving feat achieved by ChatGPT by using a protocol involving different prover and verifier instances of the gpt-5 model working collaboratively. To make sure that the produced proofs do not suffer from hallucinations, the final proof is formally verified by the lean proof assistant, and the conformance of premises and conclusion of the lean code is verified by a human. Our methodology is by no means complete or exact. It was nonetheless able to solve five out of six 2025 IMO problems, and close about a third of the sixty-six number theory conjectures in [Cohen, Journal of Integer Sequences, 2025].

Simon on Twitter / X
Can AI help connect theorems humans write in papers to proofs computers can check?We just released TheoremGraph (https://t.co/PQ8FcFQGat), and I made a 3Blue1Brown style video overview of the idea.This project was my first real research experience, and it meant a lot. Start… pic.twitter.com/yMUA0QziXM— Simon (@waskaja) June 29, 2026
Formalizing Fermat's Last Theorem
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.

Verification in the Age of AI Scientists
AI Scientists now operate at a scale that outpaces human capacity for manual review. Systems such as Sakana’s AI Scientist write entire workshop papers end-to-end. Lila Sciences runs autonomous ``AI Science Factories’’ that hypothesize, experiment, and iterate without human guidance. FutureHouse’s Kosmos and Robin generate thousands of candidate hypotheses in a single run, and Google’s Co-Scientist proposes testable experiments at a rate no laboratory can fully evaluate. Each of these systems emphasizes verified results, yet that standard ranges from near-perfect formal proof in mathematics to decade-long clinical trials in medicine, with no shared framework for judging sufficiency across domains. As outputs scale beyond what humans can manually inspect, the question of which results to trust becomes as hard as generating them. The bottleneck for AI for Science is no longer hypothesis generation, it is verification.
No. 008: AI co-mathematician, compiler-verified protocols, open infrastructure bets - The Hallway Track
The Hallway Track No. 008 - How AI is affecting the practice of science

Advancing Mathematics Research with AI-Driven Formal Proof Search
Large language models (LLMs) increasingly excel at mathematical reasoning, but their unreliability limits their utility in mathematics research. A mitigation is using LLMs to generate formal proofs in languages like Lean. We perform the first large-scale evaluation of this method's ability to solve open problems. Our most capable agent autonomously resolved 9 of 353 open Erdős problems at the per-problem cost of a few hundred dollars, proved 44/492 OEIS conjectures, and is being deployed in combinatorics, optimization, graph theory, algebraic geometry, and quantum optics research. A basic agent alternating LLM-based generation with Lean-based verification replicated the Erdős successes but proved costlier on the hardest problems. These findings demonstrate the power of AI-aided formal proof search and shed light on the agent designs that enable it.

98 out of 100. That's how many advanced maths problems current AI models managed to solve in an international benchmark in Leipzig. One of the two remaining problems? A simplified version of a… | Czech Technical University in Prague | 11 comments
98 out of 100. That's how many advanced maths problems current AI models managed to solve in an international benchmark in Leipzig. One of the two remaining problems? A simplified version of a combinatorics problem solved by researchers from the Faculty of Information Technology CTU in Prague. Alexander Clifton, Daniela Černá and Jan Volec, together with Bartołomiej Kielak from Universität Leipzig , found the solution to a simplified version of the long-standing open problem of breaking triples with permutations. The problem was later included among the 100 tasks used to test how today's AI models perform in advanced mathematics. While AI came remarkably close, this was one of the problems it couldn't solve on its own. A nice reminder that mathematical creativity and abstract reasoning still leave some room for humans. 🧠 🔗 Read the full story in the comments. | 11 comments on LinkedIn
Mathematician catches AI-generated research error by OpenAI president | Benjamin Golub posted on the topic | LinkedIn
In case you think the problem of science slop is hypothetical, here's something that happened on X yesterday: The president of OpenAI sharing wrong science slop as proud evidence of the power of mathematical AI. Within a day, the mathematician Tony Feng found a gap that undercut the paper's central argument; the author conceded, and (to his credit!) Brockman deleted the tweet as soon as he became aware. Feng was a hero to catch this. The author of the refuted claim is a mathematician but a self-described outsider to this area, with a prior big claim posted and quickly withdrawn. Greg Brockman is not an expert, either. Neither could check what they were amplifying. Heroes catching errors one at a time won't hold against the volume of AI-generated research now coming. We need an automated, independent verification layer for knowledge work, which is what we're building at Refine (refine.ink). Passing such verification should become a necessary step before big claims, especially outside one's own expertise. A link to a bigger-picture post on this is in the comments. | 23 comments on LinkedIn
Theory Beyond Theorems and Proofs: A Guest Post
Scott’s foreword: I’m extremely grateful to my brilliant colleagues, Pravesh Kothari, Raghu Meka, and Prasad Raghavendra, for sharing the guest post below about how theoretical computer…
Formalizing 100 Theorems
There is a "top 100" of mathematical theorems on the web, which is a rather arbitrary list (and most of the theorems seem rather elementary), but still is nice to look at. On the current page I keep track of which theorems from this list have been formalized. Currently the fraction that already has been formalized seems to be
The Case Against Formal Verification, 50 Years Later - Ivan Gavran
Writings on software correctness, AI, formal verification, and other technical topics.

Experts Argue Whether Computers Could Reason, and if They Should (Published 1977)
Computer world is in midst of fundamental dispute over question of computer intelligence since MIT Prof Joseph Weizenbaum wrote book arguing that machines can never be made to reason like people and should not be; Weizenbaum por (M)
VibeMathed - Math Problems Solved by AI
A community-curated record of math problems no human had solved before, proved or disproved with AI in the loop - with checkable sources and verification labels.