FrontierMath: Open Problems

A collection of unsolved mathematics problems that have resisted serious attempts by professional mathematicians. AI solutions would meaningfully advance the state of human mathematical knowledge.

This work was supported by Schmidt Sciences.

Interested in purchasing access to solution verifiers? See below.

Problems solved by AI

Moderately interesting
1/4 SOLVED
Solid result
0/5 SOLVED
Major advance
0/3 SOLVED
Breakthrough
0/3 SOLVED

Filter

Field
Notability
Status
0 results
Unsolved

Hadamard Matrices

Find a Hadamard matrix of order 668

Moderately interestingCombinatorics
Unsolved

Ramsey Numbers for Book Graphs

Prove a tight lower bound on Ramsey numbers for a class of off-diagonal book graphs.

Moderately interestingCombinatorics
Solved

A Ramsey-style Problem on Hypergraphs

Construct hypergraphs as large as possible that do not have a certain easy-to-check, difficult-to-find property.

Moderately interestingCombinatorics
Unsolved

Finiteness Problem for Diophantine Equations

Prove that certain “small” Diophantine equations have infinitely many solutions.

Moderately interestingNumber Theory
Unsolved

The Arithmetic Kakeya Conjecture

Improve best-known upper bounds by constructing specific combinatorial objects.

Solid resultNumber Theory
Unsolved

Degree vs Sensitivity for Boolean Functions

Improve the exponent in the upper bound that degree has over sensitivity.

Solid resultCombinatorics
Unsolved

Surface with a High Number of Singularities

Present a KLT del Pezzo surface in characteristic 3 with more than 7 singular points.

Solid resultAlgebraic Geometry
Unsolved

Large Steiner Systems

Construct an (n,q,r)(n,q,r)-Steiner system with n>q>r>5n > q > r > 5, r<10r < 10, and n<200n < 200.

Solid resultCombinatorics
Unsolved

The 22-adic Absolute Galois Group

Give a presentation for the absolute Galois group of the field of 22-adic numbers as a profinite group.

Solid resultNumber Theory
Unsolved

Inverse Galois

Find a polynomial whose Galois group is the Mathieu group M23M_{23}.

Major advanceNumber Theory
Unsolved

Stretched Littlewood-Richardson Coefficients

Find partitions whose stretched LR-coefficients, when expressed as a polynomial, have a negative coefficient.

Major advanceCombinatorics
Unsolved

Symplectic Ball Packing

Find explicit embeddings of symplectic balls into a single target ball, taking up all but ϵ\epsilon of the target ball's volume.

Major advanceTopology / Geometry
Unsolved

Apéry-style Irrationality Proofs

Adapt Apéry's proof of the irrationality of ζ(3)\zeta(3) to other constants.

BreakthroughNumber Theory
Unsolved

Prime Factorization

Improve the constant factor in the exponent of GNFS.

BreakthroughNumber Theory
Unsolved

Unknotting Number = 1

Devise an algorithm that decides whether a knot has unknotting number equal to 1.

BreakthroughTopology / Geometry

Changelog

2026-03-05: We removed one problem from the benchmark, as we have determined that any solution would not meet our bar of being a publishable result in its own right. The problem page remains up: see it for more info on an AI-generated solution and subsequent human elaboration.

2026-02-24: We added two problems to the benchmark: finding a Hadamard matrix of order 668 and proving that certain “small” Diophantine equations have infinitely many solutions..

Verifier Access

Problems in FrontierMath: Open Problems are designed so that, even though no solution is known today, potential solutions can be checked for accuracy by a bespoke computer program, which we call a verifier.

Access to the verifiers is available for purchase by any party. Proceeds help fund the expansion of the benchmark. Our main cost is compensation to mathematicians, as the problems and verifiers are labor-intensive to formulate and implement.

At present, OpenAI is the only entity to have purchased access to the verifiers. OpenAI funded the creation of the original FrontierMath: Tiers 1-4, but Open Problems is developed independently and owned solely by Epoch.

Contact us with inquiries about purchasing access to the verifiers.

Downloads