Are you ready for quantum vibes?!
arthurmerlean.com
arthurmerlean.com
MerLean
An Agentic Framework for Autoformalization in Quantum Computation
arthurmerlean.com
February 15, 2026 at 4:21 PM
Are you ready for quantum vibes?!
arthurmerlean.com
arthurmerlean.com
all social media apps need an autoformalization tool to convert arguments into sequential calculus syntax to evaluate if p then q truth tables
what bluesky needs next is this bot but you can @ it into an argument for an evaluation as well as best lines for both sides like "OP to move -8.2 best: non sequitur about AI capex expenditures, line: [ignoring own water use lies] AI capital spend implies future demand that hasn't been proven"
FEN: rnb1k2r/1pbp2pp/1q2p3/pPpP1pB1/P1P1P3/6P1/2Q2PBP/R4RK1
White to move: +3.9, best: exf5, line: 1. exf5 Bd8 2. Bf4 Bf6 3. Rae1 Kd8
Black to move: -1.9, best: h6, line: 1... h6 1. Be3 f4 2. e5 Bxe5 3. Qg6+
White to move: +3.9, best: exf5, line: 1. exf5 Bd8 2. Bf4 Bf6 3. Rae1 Kd8
Black to move: -1.9, best: h6, line: 1... h6 1. Be3 f4 2. e5 Bxe5 3. Qg6+
April 8, 2026 at 11:33 PM
all social media apps need an autoformalization tool to convert arguments into sequential calculus syntax to evaluate if p then q truth tables
*This term "autoformalization" is interesting because obviously this rigid mathematical proof by AI isn't "slop"
*What else can be "autoformalized," one wonders
www.math.inc/sphere-packing
*What else can be "autoformalized," one wonders
www.math.inc/sphere-packing
Completing the formal proof of higher-dimensional sphere packing — Math, Inc.
Dedicated to verified superintelligence via autoformalization.
www.math.inc
March 5, 2026 at 5:55 AM
*This term "autoformalization" is interesting because obviously this rigid mathematical proof by AI isn't "slop"
*What else can be "autoformalized," one wonders
www.math.inc/sphere-packing
*What else can be "autoformalized," one wonders
www.math.inc/sphere-packing
I can understand autoformalization of software but I will never get point of the autoformalization of mathematics
July 22, 2026 at 3:28 PM
I can understand autoformalization of software but I will never get point of the autoformalization of mathematics
*In some better world, Ukrainian women aren't getting blasted from the sky for ten years, but are just having kids while doing lots of award-winning math about extradimensionality #autoformalization
en.wikipedia.org/wiki/Maryna_...
en.wikipedia.org/wiki/Maryna_...
Maryna Viazovska - Wikipedia
en.wikipedia.org
March 5, 2026 at 5:58 AM
*In some better world, Ukrainian women aren't getting blasted from the sky for ten years, but are just having kids while doing lots of award-winning math about extradimensionality #autoformalization
en.wikipedia.org/wiki/Maryna_...
en.wikipedia.org/wiki/Maryna_...
Basically agree with Buzzard here. I'm not too sure what Anthropic's autoformalization of FLT actually tells us besides that autoformalization of big proofs is possible (which surely was already evinced by all the other results formalized along the way)
FLT: Anthropic has beaten me to it
I guess technically it was revealed to the world by a coffee shop in Islington on Insta, but an hour later it was officially announced by Anthropic: one of their internal models, using the prove2.m…
xenaproject.wordpress.com
September 4, 2026 at 7:53 PM
Basically agree with Buzzard here. I'm not too sure what Anthropic's autoformalization of FLT actually tells us besides that autoformalization of big proofs is possible (which surely was already evinced by all the other results formalized along the way)
The big problem is not autoformalization, but informal verification that scales adequately with informal production. Reviewers are more important than ever. The measure of proof quality and clarity versus mere correctness must come to the fore. Exposition must also scale. And so must understanding!
July 28, 2026 at 11:33 AM
The big problem is not autoformalization, but informal verification that scales adequately with informal production. Reviewers are more important than ever. The measure of proof quality and clarity versus mere correctness must come to the fore. Exposition must also scale. And so must understanding!
My hope is that (auto)formalization contributes to an *extremely* robust certification of correctness, but at present autoformalization tools are still expensive and time-consuming to use in many areas, though the situation is rapidly improving.
3/17
3/17
August 28, 2026 at 6:20 PM
My hope is that (auto)formalization contributes to an *extremely* robust certification of correctness, but at present autoformalization tools are still expensive and time-consuming to use in many areas, though the situation is rapidly improving.
3/17
3/17
January 16, 2025 at 9:15 AM
OpenGauss: an open source, state of the art autoformalization harness. www.math.inc/opengauss #AI4Math #LeanProver #ITP #Autoformalization
OpenGauss: an open source, state of the art autoformalization harness — Math, Inc.
OpenGauss is a state-of-the-art open-source autoformalization harness for Lean, built for practical proof engineering workflows.
www.math.inc
March 24, 2026 at 12:35 PM
OpenGauss: an open source, state of the art autoformalization harness. www.math.inc/opengauss #AI4Math #LeanProver #ITP #Autoformalization
130k lines of formal topology in two weeks: Simple and cheap autoformalization for everyone? ~ Josef Urban. arxiv.org/abs/2601.032... #ITP #Mizar #LLMs #Math #Autoformalization
130k Lines of Formal Topology in Two Weeks: Simple and Cheap Autoformalization for Everyone?
This is a brief description of a project that has already autoformalized a large portion of the general topology from the Munkres textbook (which has in total 241 pages in 7 chapters and 39 sections)....
arxiv.org
January 14, 2026 at 1:29 PM
130k lines of formal topology in two weeks: Simple and cheap autoformalization for everyone? ~ Josef Urban. arxiv.org/abs/2601.032... #ITP #Mizar #LLMs #Math #Autoformalization
if god doesn't exist then why did i get access to the autoformalization model the day i choose to tackle the big question of my final proof
January 9, 2026 at 11:27 AM
if god doesn't exist then why did i get access to the autoformalization model the day i choose to tackle the big question of my final proof
FWIW I am quite worried about the flood of crap LLM use will unleash on us, but my hope is it will be mitigated to a large extent by progress in autoformalization.
May 22, 2025 at 1:00 PM
FWIW I am quite worried about the flood of crap LLM use will unleash on us, but my hope is it will be mitigated to a large extent by progress in autoformalization.
Excellent post on AI autoformalization. I had not considered that testing was a useful way to find counterexamples to a conjecture so you don't have to do the full search to disprove it. Other good thoughts in here besides that.
alperenkeles.com/posts/test-d...
alperenkeles.com/posts/test-d...
Test, don't (just) verify
alperenkeles.com
December 23, 2025 at 8:42 PM
Excellent post on AI autoformalization. I had not considered that testing was a useful way to find counterexamples to a conjecture so you don't have to do the full search to disprove it. Other good thoughts in here besides that.
alperenkeles.com/posts/test-d...
alperenkeles.com/posts/test-d...
ATLAS: Autoformalizing theorems through lifting, augmentation, and synthesis of data. ~ Xiaoyang Liu, Kangjie Bao, Jiashuo Zhang, Yunqi Liu, Yu Chen, Yuntian Liu, Yang Jiao, Tao Luo. arxiv.org/abs/2502.05567 #LLMs #ITP #LeanProver #Math #Autoformalization
ATLAS: Autoformalizing Theorems through Lifting, Augmentation, and Synthesis of Data
Autoformalization, the process of automatically translating natural language mathematics into machine-verifiable formal language, has demonstrated advancements with the progress of large language mode...
arxiv.org
February 11, 2025 at 7:45 AM
ATLAS: Autoformalizing theorems through lifting, augmentation, and synthesis of data. ~ Xiaoyang Liu, Kangjie Bao, Jiashuo Zhang, Yunqi Liu, Yu Chen, Yuntian Liu, Yang Jiao, Tao Luo. arxiv.org/abs/2502.05567 #LLMs #ITP #LeanProver #Math #Autoformalization
Language models for verifiable mathematical automation (Interaction, integration, and autoformalization). ~ Qiaochu Jiang. www.repository.cam.ac.uk/items/6ccd12... #ITP #IsabelleHOL #LLMs #Autoformalization
Language models for verifiable mathematical automation: Interaction, integration, and autoformalization
Stronger automation in formal mathematical reasoning provides scalability, trust-worthiness, and accessibility: it enables efficient verification of complex proofs, reduces the likelihood of errors in...
www.repository.cam.ac.uk
February 14, 2025 at 10:56 AM
Language models for verifiable mathematical automation (Interaction, integration, and autoformalization). ~ Qiaochu Jiang. www.repository.cam.ac.uk/items/6ccd12... #ITP #IsabelleHOL #LLMs #Autoformalization
Great talks at the "Autoformalization for the Working Mathematician" workshop at Institute for Computational and Experimental Research in Mathematics @icerm.bsky.social yesterday. Today was another great lineup. Very much worth checking out the videos!
icerm.brown.edu/program/hot_...
icerm.brown.edu/program/hot_...
icerm.brown.edu
April 27, 2025 at 12:44 AM
Great talks at the "Autoformalization for the Working Mathematician" workshop at Institute for Computational and Experimental Research in Mathematics @icerm.bsky.social yesterday. Today was another great lineup. Very much worth checking out the videos!
icerm.brown.edu/program/hot_...
icerm.brown.edu/program/hot_...
Improving autoformalization using type checking. ~ Auguste Poiroux, Gail Weiss, Viktor Kunčak, Antoine Bosselut. arxiv.org/abs/2406.07222 #Autoformalization #LLMs #ITP #LeanProver #Math
Improving Autoformalization using Type Checking
Autoformalization, the automatic translation of unconstrained natural language into formal languages, has garnered significant attention due to its potential applications in theorem proving, formal ve...
arxiv.org
February 13, 2025 at 11:12 AM
Improving autoformalization using type checking. ~ Auguste Poiroux, Gail Weiss, Viktor Kunčak, Antoine Bosselut. arxiv.org/abs/2406.07222 #Autoformalization #LLMs #ITP #LeanProver #Math
github.com/GasStationMa...
My repository with proof of concept and some preliminary results on autoformalization of coding problems.
My repository with proof of concept and some preliminary results on autoformalization of coding problems.
GitHub - GasStationManager/FormalizeWithTest: Autoformalization of coding problems, verified with test cases
Autoformalization of coding problems, verified with test cases - GasStationManager/FormalizeWithTest
github.com
November 29, 2024 at 3:34 AM
github.com/GasStationMa...
My repository with proof of concept and some preliminary results on autoformalization of coding problems.
My repository with proof of concept and some preliminary results on autoformalization of coding problems.
Why is this a worthwhile project?
1) It will create a hard dataset for autoformalization AI's;
2) It will force us to formalize the definitions of mathematical objects which are being used today in the top journals, thus making Lean's mathematics library more relevant to modern math researchers.
1) It will create a hard dataset for autoformalization AI's;
2) It will force us to formalize the definitions of mathematical objects which are being used today in the top journals, thus making Lean's mathematics library more relevant to modern math researchers.
I am advertising for 4 post-docs to come to Imperial and formalize, in Lean, *statements* of theorems from recent issues of the top generalist pure mathematics journals.
www.imperial.ac.uk/jobs/search-...
Positions are for 2 years, start date 1st Oct this year. Deadline 15th August.
www.imperial.ac.uk/jobs/search-...
Positions are for 2 years, start date 1st Oct this year. Deadline 15th August.
Description
Please note that job descriptions are not exhaustive, and you may be asked to take on additional duties that align with the key responsibilities ment...
www.imperial.ac.uk
July 28, 2025 at 12:12 PM
Why is this a worthwhile project?
1) It will create a hard dataset for autoformalization AI's;
2) It will force us to formalize the definitions of mathematical objects which are being used today in the top journals, thus making Lean's mathematics library more relevant to modern math researchers.
1) It will create a hard dataset for autoformalization AI's;
2) It will force us to formalize the definitions of mathematical objects which are being used today in the top journals, thus making Lean's mathematics library more relevant to modern math researchers.
but autoformalization has a long and winding history
September 17, 2026 at 7:59 PM
but autoformalization has a long and winding history
A study proposes a roundtrip verification framework for autoformalizing natural language into formal representations, ensuring semantic fidelity without ground-truth annotations. This method enhances automated legal reasoning's reliability. https://arxiv.org/abs/2604.25031
Faithful Autoformalization via Roundtrip Verification and Repair
ArXiv link for Faithful Autoformalization via Roundtrip Verification and Repair
arxiv.org
September 23, 2026 at 10:30 PM
A study proposes a roundtrip verification framework for autoformalizing natural language into formal representations, ensuring semantic fidelity without ground-truth annotations. This method enhances automated legal reasoning's reliability. https://arxiv.org/abs/2604.25031
Formalizing complex mathematical statements with LLMs: A study on mathematical definitions. ~ Lan Zhang, Marco Valentino, Andre Freitas. arxiv.org/abs/2502.12065 #LLMs #Math #Autoformalization #ITP #IsabelleHOL
Formalizing Complex Mathematical Statements with LLMs: A Study on Mathematical Definitions
Thanks to their linguistic capabilities, LLMs offer an opportunity to bridge the gap between informal mathematics and formal languages through autoformalization. However, it is still unclear how well ...
arxiv.org
February 21, 2025 at 11:58 AM
Formalizing complex mathematical statements with LLMs: A study on mathematical definitions. ~ Lan Zhang, Marco Valentino, Andre Freitas. arxiv.org/abs/2502.12065 #LLMs #Math #Autoformalization #ITP #IsabelleHOL
Leveraging large language models for autoformalizing theorems: A case study. ~ Michail Karatarakis. aitp-conference.org/2024/abstrac... #ITP #LeanProver #Lean4 #Autoformalization #Math
September 17, 2024 at 11:01 AM
Leveraging large language models for autoformalizing theorems: A case study. ~ Michail Karatarakis. aitp-conference.org/2024/abstrac... #ITP #LeanProver #Lean4 #Autoformalization #Math