Lean is a dependently-typed programming language and theorem prover.
Seattle
Joined April 2018
- Tweets882
- Following52
- Followers14.8K
- Likes816
๐๐๐๐ง ๐
๐๐ ๐ข๐ฌ ๐ก๐ข๐ซ๐ข๐ง๐ ๐ ๐๐๐๐ก๐ง๐ข๐๐๐ฅ ๐๐จ๐ฆ๐ฆ๐ฎ๐ง๐ข๐๐๐ญ๐ข๐จ๐ง๐ฌ & ๐๐จ๐ง๐ญ๐๐ง๐ญ ๐๐๐๐
We are looking for a hands-on communicator and content creator to own Lean FROโs external communications. The role is full-time and remote, working directly with Lean developers and leadership.
We expect the successful candidate to turn technical developments into clear, engaging stories, build a consistent public presence, and help audiences understand Lean accurately. They do not need to be a Lean expert when they join, but must be eager and able to learn enough to communicate accurately.
This is a chance to turn substantive technical work into content people understand, and to take ownership of a communications function whose direction you can help shape.
๐
๐ฎ๐ฅ๐ฅ ๐ฃ๐จ๐ ๐ฉ๐จ๐ฌ๐ญ๐ข๐ง๐ ๐๐ง๐ ๐๐ฉ๐ฉ๐ฅ๐ข๐๐๐ญ๐ข๐จ๐ง: jobs.lever.co/convergentreseโฆ
#LeanLang #LeanProver
Lean retweeted
Thrilled to announce that our @GoogleDeepMind team has cracked a major challenge in software engineering and cybersecurity: formally verifying a real-world implementation of the #Signal cryptographic protocol at the machine-code level. This proof was done using AlphaProof Code, our new AI framework for formal specification and verification.
* Paper: deepmind.google/documents/36โฆ
* Lean code: github.com/google-deepmind/cโฆ
* DeepMind blog post: deepmind.google/blog/introduโฆ
A huge shoutout to my amazing teammates: @tchajed,
@FerrousAndrew, Jonathan Katz, @pushmeet, @antonkovsharov, Karn Seth, Sam Schlesinger, @RandomlyWalking.
I started working on formal software verification 20+ years ago as a grad student. I was persuaded by Dijkstra and other computer science pioneersโ vision of programming backed up by mathematical theorems. At that time, we had to choose between automated code analysis algorithms that scaled but could only prove shallow properties, and deep manual verification that took decades of expert human work.
Thanks to progress in AI for mathematical reasoning, we donโt have to make that choice any more. Over four weeks in October 2026, AlphaProof Code agents generated more than 6 million lines of machine-checked Lean proof. We proved deep computational and symbolic security properties of the entire protocol core of Signal, all the way down to the ARM machine-code level.
This is just the first step, of course, because our goal is nothing short of verifying the worldโs software infrastructure. Most cyberattacks, including the AI-powered ones recently in the news, happen because of preventable code bugs. I expect that agents for formal code analysis will eliminate these bugs one category at a time.
As our paper explains in more detail, this doesnโt at all mean humans are out of the picture! Proofs are now largely automated, but human experts will need to decide what is proved with AI assistance. AlphaProof Code offers an early glimpse of what such human-AI collaboration could look like, but many other proposals of this nature will emerge.
We have released all our Lean specs and proofs and plan to work with the open-source community on improving them. If you are building critical software and would like to apply these ideas, come talk to us!
Lean retweeted
I had Opus 5.5 formalize CPython in Lean: 613,318 lines of Lean that copy Python's internals line by line, tested against the real interpreter and checked by Lean.
On the way it found real bugs in Python, including crashes from plain Python code. ๐งต
Software Verification in Lean is live now! You can still register via Luma to access the livestream link: luma.com/kytl0zlj
We're very excited for ๐ฆ๐ผ๐ณ๐๐๐ฎ๐ฟ๐ฒ ๐ฉ๐ฒ๐ฟ๐ถ๐ณ๐ถ๐ฐ๐ฎ๐๐ถ๐ผ๐ป ๐ถ๐ป ๐๐ฒ๐ฎ๐ป tomorrow! The in-person event is full, but livestream spots remain open: Register on Luma to receive the link: luma.com/kytl0zlj
๐ก๐ฒ๐๐น๐ ๐ฎ๐ฑ๐ฑ๐ฒ๐ฑ ๐๐ผ ๐๐ต๐ฒ ๐ฝ๐ฟ๐ผ๐ด๐ฟ๐ฎ๐บ: A closing round-table discussion on AI safety and formal verification (4:45โ5:30 pm PT), moderated by Pedro Abreu (@p_droabreu0), host of the Type Theory Forall podcast, featuring Max Tegmark, Leonardo de Moura, Jonathan Protzenko, Andres Erbsen and Mike Dodds.
๐ See the event website for the full program: beneficial-ai-foundation.gitโฆ
We're very excited for ๐ฆ๐ผ๐ณ๐๐๐ฎ๐ฟ๐ฒ ๐ฉ๐ฒ๐ฟ๐ถ๐ณ๐ถ๐ฐ๐ฎ๐๐ถ๐ผ๐ป ๐ถ๐ป ๐๐ฒ๐ฎ๐ป tomorrow! The in-person event is full, but livestream spots remain open: Register on Luma to receive the link: luma.com/kytl0zlj
๐ก๐ฒ๐๐น๐ ๐ฎ๐ฑ๐ฑ๐ฒ๐ฑ ๐๐ผ ๐๐ต๐ฒ ๐ฝ๐ฟ๐ผ๐ด๐ฟ๐ฎ๐บ: A closing round-table discussion on AI safety and formal verification (4:45โ5:30 pm PT), moderated by Pedro Abreu (@p_droabreu0), host of the Type Theory Forall podcast, featuring Max Tegmark, Leonardo de Moura, Jonathan Protzenko, Andres Erbsen and Mike Dodds.
๐ See the event website for the full program: beneficial-ai-foundation.gitโฆ
Software Verification in Lean (SVIL) comes to Stanford University on Saturday, November 7.
Co-organized by Beneficial AI Foundation and Lean FRO, it is a day of public lectures on advancing Leanโs software verification ecosystem toward production use, with a focus on end-to-end verification of security-critical systems.
Talks will cover Rust verification with the Aeneas framework, low-level software verification, and formally verified OS kernels.
Speakers include Max Tegmark, Shaowei Lin, Leonardo de Moura, Jonathan Protzenko, Andres Erbsen, and Mike Dodds.
In-person space is limited. Non-local and waitlisted registrants will receive a livestream link, and recordings will be available afterward.
Program and registration: beneficial-ai-foundation.gitโฆ
Lean retweeted
Most think vibecoding at scale is slop. Hereโs a counterexample.
Today we release a 16M-line formalization of the entire Stacks Project, critical infra underlying modern math, with optimized code quality and its highly interconnected structure preserved.
A mathematical city ๐
The #TauCeti project is using AI to build a library of formalized mathematics: Human-written roadmaps specify the intended mathematics, and the #Lean kernel checks every proof the AI contributors produce.
Adversarial AI reviewers then assess whether the results match what the roadmaps intend. Kim Morrison writes that this process โbrings the material up to a standard we can keep building on.โ
The aim: reusable formalizations of โall of basic mathematics.โ The project is incubated by Lean FRO and the Mathlib Initiative, and open to contributions from everyone.
See the Tau Ceti use case for more: lean-lang.org/use-cases/tau-โฆ
Oh wow, we love this. Harsh Gupta's Lean 4 submission to the "$12 server challenge" proves that, under certain conditions, posts are โalready stored in the exact order they need,โ so it can skip the extra work and read them directly.
And bypassing SQLiteโs normal query engine cut the time to load the feed from 16.7 ms to 5.4 ms.
Congrats @hargup13!
It's official now, my Lean 4 submission is #1 on "$12 server challenge", serving more users than Rust, C or even hand written assembly.
The creator of a the challenge did a pretty good job explaining how Lean 4 allows you to create software even more optimized than handwritten C.
Lean retweeted
My first book, The Proof in the Code, has been named one of the best nonfiction books of 2026 by @PublishersWkly
"Itโs one thing to make complex mathematical concepts accessible to a lay audience, but to illustrate their importance through an engrossing, character-driven narrative is a feat only a journalist as adept as Hartnett can pull off. This cinematic account of the creation of Lean, a software tool thatโs become vital for mathematicians and AI researchers, is necessary reading for understanding the rise of human-computer collaboration."
Lean retweeted
We're in @nytimes: if AI can prove open math problems, can it prove software has no bugs? That is what we work on, with Lean as the foundation for machine-checked trust.
nytimes.com/2026/10/29/scienโฆ
Lean retweeted
I'm thrilled to bring &Truth out of stealth as its co-founder and Chief Science Officer. With a team I'm proud to be part of, we're setting out to specify and prove all the world's software correct and secure.
If you geek out over code people can understand and trust, join us!
We are out of stealth! As AI writes more code, we're building tools to prove programs meet precise requirements, with Lean proofs anyone can independently check. Let's make verified software the norm in every language and application.
Come take a look: andtruth.com
Lean retweeted
We are out of stealth! As AI writes more code, we're building tools to prove programs meet precise requirements, with Lean proofs anyone can independently check. Let's make verified software the norm in every language and application.
Come take a look: andtruth.com
Highlighting two great sites for the latest on Lean, both built and curated by @soonhokong .
๐ ๐๐๐ฉ๐๐ซ๐ฌ ๐๐ข๐ญ๐ก ๐๐๐๐ง: "arXiv papers that use Lean, the interactive theorem prover. Updated daily." You can browse by topic or by the last 7 or 30 days.
paperswithlean.com
๐ฐ ๐๐๐๐ง ๐๐๐๐ค๐ฅ๐ฒ: News about Lean, from Lean core to verification to AI and formalization, organized by week.
leanweekly.news
We're grateful to Soonho for this work.
Software Verification in Lean on November 7 has reached in-person capacity! If you're interested in attending live, please join the waitlist: A livestream link will be shared with all waitlisted registrants a few days before the event.
Software Verification in Lean (SVIL) comes to Stanford University on Saturday, November 7.
Co-organized by Beneficial AI Foundation and Lean FRO, it is a day of public lectures on advancing Leanโs software verification ecosystem toward production use, with a focus on end-to-end verification of security-critical systems.
Talks will cover Rust verification with the Aeneas framework, low-level software verification, and formally verified OS kernels.
Speakers include Max Tegmark, Shaowei Lin, Leonardo de Moura, Jonathan Protzenko, Andres Erbsen, and Mike Dodds.
In-person space is limited. Non-local and waitlisted registrants will receive a livestream link, and recordings will be available afterward.
Program and registration: beneficial-ai-foundation.gitโฆ
Lean retweeted
We also thank the Lean and Mathlib communities.
Completing the formalization is one milestone; making it more useful to mathematicians is ongoing work.
Weโll maintain and improve this codebase: faster builds, clearer organization, and results that are easier to find and reuse.
Congratulations to the Axiom team on this expansive undertaking.
Lean retweeted
1/๐งต Coding agents can make code faster. Can they also prove it still does the same thing?
Introducing Agentic Verified Optimization ๐ Agents return faster code plus a Lean proof that it equals the original on every input.
๐ Blog: aqueduct.verina.io
๐ Paper: arxiv.org/abs/2610.30188
Lean retweeted
9/ ๐ Paper: arxiv.org/abs/2610.30188
๐ Blog: aqueduct.verina.io
Amazing work by @0xlf_, @udayaghai, @uniq_zz, @dawnsongtweets, @soonhokong ๐
This work is done during my internship at @AmazonScience.
Software Verification in Lean (SVIL) comes to Stanford University on Saturday, November 7.
Co-organized by Beneficial AI Foundation and Lean FRO, it is a day of public lectures on advancing Leanโs software verification ecosystem toward production use, with a focus on end-to-end verification of security-critical systems.
Talks will cover Rust verification with the Aeneas framework, low-level software verification, and formally verified OS kernels.
Speakers include Max Tegmark, Shaowei Lin, Leonardo de Moura, Jonathan Protzenko, Andres Erbsen, and Mike Dodds.
In-person space is limited. Non-local and waitlisted registrants will receive a livestream link, and recordings will be available afterward.
Program and registration: beneficial-ai-foundation.gitโฆ
Lean retweeted
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.
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.
Our system found its natural language disproof for the ETH we reached out to to @rrwilliams. Ryan will form a collaboration with Russell Impagliazzo and Ramamohan Paturi to verify and shepherd this result. Impagliazzo and Paturi, along with Francis Zane, are the original proposers of the Exponential Time Hypothesis, and Williams has refuted strong versions of the ETH, used ETH for hardness results, and connected algorithms and lower bounds.
This group asked us to make all raw data of the results public to facilitate independent verification by the community, which we have done by posting the natural language proof and Lean formalization to our GitHub repo. The group also 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.
github.com/openai/math/blob/โฆ
