About FrontierMath: Open Problems
Tracks whether AI systems can solve selected open problems and advance mathematical knowledge.
Key differences
Compare each service by the same aspect.
| Aspect | AI Proof Watch | FrontierMath: Open Problems |
|---|---|---|
| Subject coverage | Mathematics, theoretical computer science, and theoretical physics. | Mathematics, with a constructive and computationally verifiable bias. |
| Main purpose | Curate reported contributions and make claims, sources, and evidence easier to inspect; list linked human follow-up works. | Tracks whether AI systems can solve selected open problems and advance mathematical knowledge. |
| Assessment method | Four independent ratings cover AI involvement, technical verification, human understanding, and importance. Importance assumes the claim is correct; it is not a verification score. | Working mathematicians propose problems and an editorial board assigns notability tiers. Custom verifiers check candidate solutions; a positive verifier result can be evidence rather than a full proof. Solved problems are retired. |
| Email updates | Weekly updates across all topics, or selected-interest alerts that are daily by default and sent only when there is an update. | Epoch tests selected frontier models and updates individual problem records. We found no email subscription on the reviewed benchmark pages. |
When to use each
Use FrontierMath: Open Problems to evaluate AI on a stable, curated set of open problems whose candidate solutions can be checked computationally.
By comparison, AI Proof Watch records reported contributions, including work that is not amenable to automatic verification, across mathematics, theoretical computer science, and theoretical physics. It rates technical verification separately from importance and AI involvement.
Sources
Feature comparison checked 20 October 2026. Public pages may change.