Updates on my research and expository papers, discussion of open problems, and other maths-related topics. By Terence Tao
oai-math-revised: a community hub for revising OpenAI’s math write-ups
24 October, 2026 in advertising | Tags: Abhinav Nellore , Johannes Schmitt , Rachel Ward , Soledad Villar | by Terence Tao
[This is a guest post by Abhinav Nellore , Johannes Schmitt , Soledad Villar , and Rachel Ward – T.]
OpenAI’s math write-ups at https://github.com/openai/math can be difficult to understand and can have incomplete citations.
https://github.com/mathcommunity/oai-math-revised/ seeks to be an inclusive effort to improve these write-ups, and the broader math community–students, postdocs, professors, and enthusiasts–is invited to participate.
The repo is browsable at https://mathcommunity.github.io/oai-math-revised/
We also invite the community to email mathcommunityadmin@gmail.com if they want to serve as admins/reviewers and for which categories from https://github.com/openai/math/blob/main/overview.pdf they’d be appropriate for reviews.
Share this:
Print (Opens in new window) Print
Email a link to a friend (Opens in new window) Email
Share on WhatsApp (Opens in new window) WhatsApp
Share on Bluesky (Opens in new window) Bluesky
Share on Nextdoor (Opens in new window) Nextdoor
Share on Threads (Opens in new window) Threads
Share on Tumblr (Opens in new window) Tumblr
Share on Telegram (Opens in new window) Telegram
Share on LinkedIn (Opens in new window) LinkedIn
Share on X (Opens in new window) X
Share on Facebook (Opens in new window) Facebook
Share on Reddit (Opens in new window) Reddit
Share on Mastodon (Opens in new window) Mastodon
Share on Pinterest (Opens in new window) Pinterest
Recent Comments
Top Posts
Mathematics is a subject for the young
oai-math-revised: a community hub for revising OpenAI’s math write-ups
Why are we even doing all this?
Simple and critical zeros of the Riemann zeta-function
IAS applications for membership in the school of mathematics
A Severe Misalignment of AI in Mathematics
A sketch of the proof of non-existence of Siegel zeroes
Will it be a golden age or a dark age for mathematics (and science)?
An experimental AI-math challenge: using tenth discriminant bounds to explore arguments ruling out Siegel zeroes
Archives
Categories
non-technical (239) admin (49)
teaching (192) 245A – Real analysis (11)
247A – Classical Fourier Analysis (2)
247B – Classical Fourier Analysis (5)
254A – analytic prime number theory (19)
254A – Hilbert's fifth problem (12)
254A – Incompressible fluid equations (5)
254B – expansion in groups (8)
254B – Higher order Fourier analysis (9)
255B – incompressible Euler equations (2)
285G – poincare conjecture (20)
The Polymath Blog
A sort of Polymath on a famous MathOverflow problem
Polymath proposal: finding simpler unit distance graphs of chromatic number 5
A new polymath proposal (related to the Riemann Hypothesis) over Tao’s blog
Spontaneous Polymath 14 – A success!
Non-transitive Dice over Gowers’s Blog
9 comments
So it is up to unpaid volunteers to sift through OpenAI’s slop and refine their output for them? After all the fanciful debate about putting pressure on these trillion-dollar companies, nothing has fundamentally shifted in our approach?
It’s good to know that you can always count on academia to never change for the better even in the face of the most extreme circumstances.
Why do you think it’s “for them” and not for the community itself?
Given the nature of the existing tension between the mathematical community and frontier labs, I have to wonder, is this really the right thing to do? To engage with the labs in good faith, cleaning up and polishing their results for them?
It seems that this will mainly serve to validate the approach that OpenAI has taken, and to pigeonhole mathematicians into the role of interpreting the “gifts” the “machine gods” have bestowed upon us, further distracting them from what should be their main purpose right now, which is figuring out how to act autonomously as a community going forward.
In addition, these improved writeups will be used as training data for the future models. OpenAI will eat these up and spit you out without so much as a thank you.
Several of the authors appear to be heavily involved in machine learning and AI for math. While it is expected that initiatives would come from such a direction, it is not really a neutral organization.
The repository is separate from OpenAI, but participants will work for free for OpenAI.
Why is it our job to revise OpenAI’s manuscripts? They are the authors and they are responsible for the content and the quality of the writing.
We are interested in understanding the scientific advance, regardless of the source, and we are fully within our rights to produce review articles and share our understanding, but it seems weird for me to spend volunteer efforts into improving write-ups that belong to a wealthy for-profit company. If they want these services they should pay for them.
Really strange. Why would I revise somebody else’s manuscript? I would use the result (or idea) and cite it as it is.
What OpenAI did to math is akin to somebody dropping a gigantic database of environmental DNA ( https://en.wikipedia.org/wiki/Environmental_DNA ) without really explaining the provenance or anything about it, in order to show the world they have the biggest, baddest DNA sequencing machine. This is quite different from actually writing 300 biology papers.
Now biologists will wish to use that database for interesting stuff of course! Including crowd-sourcing efforts to understand what’s in there. But claiming that we are cleaning up 100s of papers of the DNA dumper… the dumper just dumped a pile of stuff without any intention to meaningfully engage academically.
One could claim that the individual raw OAI dump pieces have more potential academic value than contextless DNA. I think that’s 2025 thinking
This is not “revising” as in, fixing their work for them. This is “revising” as in … a big corporation dumped a bunch of toxic sludge over a national park, and people are volunteering to help clean up the sludge because they care deeply about the long term well being of the park.
We can talk later about how said corporation will be held accountable for their toxic sludge dump
Not sure revision is the right approach to take. It makes more sense to have independent papers on the same results, and have the website serve as a reference to these papers.
Leave a Reply Cancel reply
For commenters
To enter in LaTeX in comments, use $latex <Your LaTeX code> $ (without the < and > signs, of course; in fact, these signs should be avoided as they can cause formatting errors). Also, backslashes \ need to be doubled as \\. See the about page for details and for other commenting policy.
Blog at WordPress.com. Ben Eastaugh and Chris Sternal-Johnson.
Subscribe Subscribed What's new Have a WordPress.com account? Log in now.
What's new Copy shortlink View post in Reader Manage subscriptions Sign up Log in Report this content Collapse this bar