The goal of this repository is to collect statements of problems I (Daniel Litt) like, with some certificate of their interest, as a way to measure the progress of AI towards solutions to questions I personally care about. It will reflect my own idiosyncratic preferences and interests. I will try to add a couple such problems each week. Each problem will contain the following information:

• The problem statement
• A link to a brief blog post on the problem, indicating both what is known and some of the relevant history
• My sense of whether the problem statement is currently formalizable in Lean
• A brief comment on why I think the problem is interesting, how interesting I think it is, and whether or not it is likely to lead to new mathematics
• Some speculation on the problem's difficulty

The problems are selected by Daniel Litt, a mathematician working in algebraic geometry and number theory, broadly construed. The site itself was set up with a near-infinite amount of help from Thomas Bloom, the mathematician behind ErdosProblems.com.

The best way to do this is to add a comment under the problem.

To submit a solution or partial result, please navigate to the blog post linked in the problem statement and post a comment indicating what you've found, with a link to your writeup. If you believe that you (either alone or in concert with an AI tool) have generated a solution to one of these problems, please check it carefully before submitting it. In addition to trying to verify it on your own, audit it as follows. Upload a .pdf or TeX writeup to GPT 5.x Pro (or a model of similar strength) and prompt it with: "Please give me detailed comments on the attached, identifying any errors or points where the argument is unclear." Do not submit it unless this audit indicates the writeup is fully correct; when you submit, please link to the GPT 5.x Pro audit.

If you'd like to submit a problem, you should feel free to send me an email with your submission. In this case please follow the format outlined above, and draft a guest blog post on the problem. I don't guarantee that I will accept it, and please don't take my refusal or lack of response personally--this is meant to be an individual and idiosyncratic list of problems I care about.