By verification grade
11 of 28 carry an independent check. Open the review queue.
Open longest before falling
- 177 yrPrime gaps at most 186, conditional on three unproved Lean axiomsposed 1849 · GPT-6 Astra
- 167 yrQuasi-Riemann hypothesis: no zeros of zeta or any Dirichlet L-function with real part above 7/8posed 1859 · Unreleased internal OpenAI model, not named in the release
- 99 yrArtin's primitive root conjecture: infinitely many primes for every admissible baseposed 1927 · Unreleased internal OpenAI model, not named in the release
- 96 yrLean disproof of Krempa's matrix form of the Koethe conjectureposed 1930 · GPT-6 Astra (pre-release)
From the 17 of 28 entries recording a posed year. Add a missing one.
| Date | Finding | Lab | Model | Verification | Autonomy | Field |
|---|---|---|---|---|---|---|
| 2026-10-26 | A subexponential counting algorithm for fixed-width CNF, refuting the Exponential Time Hypothesis | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | AI-led | computer-science |
| 2026-10-26 | Sogge's local smoothing conjecture claimed in every dimension, with Bochner–Riesz and Kakeya as consequences | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | AI-led | mathematics |
| 2026-10-06 | 719 manuscripts in 372 result families from an unreleased OpenAI model | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | AI-led | mathematics |
| 2026-10-06 | Quasi-Riemann hypothesis: no zeros of zeta or any Dirichlet L-function with real part above 7/8 | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | Collaborative | mathematics |
| 2026-10-06 | Kakeya maximal conjecture in three dimensions and the Kakeya set conjecture in four | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | AI-led | mathematics |
| 2026-10-06 | Counterexample to the Borel conjecture in dimension four | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | AI-led | mathematics |
| 2026-10-06 | Artin's primitive root conjecture: infinitely many primes for every admissible base | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | AI-led | mathematics |
| 2026-10-06 | Modularity of every elliptic curve over every imaginary quadratic field | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | AI-led | mathematics |
| 2026-10-06 | Negative answer to the hyperinvariant subspace problem on Hilbert space | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | AI-led | mathematics |
| 2026-10-06 | Massive continuum limit of the two-dimensional O(3) model | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | AI-led | mathematics |
| 2026-10-06 | Proof of Barnette's conjecture on Hamiltonian cycles in cubic bipartite planar graphs | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | AI-led | mathematics |
| 2026-10-06 | Shelah's eventual categoricity conjecture for abstract elementary classes | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | AI-led | mathematics |
| 2026-10-06 | Matrix multiplication exponent bounded by 9/4 | OpenAI | Unreleased internal OpenAI model, not named in the release | Formally verified | AI-led | computer-science |
| 2026-10-06 | Integer multiplication below n log n, against the Schönhage–Strassen optimality conjecture | OpenAI | Unreleased internal OpenAI model, not named in the release | Claimed | AI-led | computer-science |
| 2026-09-08 | Finite-time blowup for Navier-Stokes with smooth forcing, Clay alternatives C and D | OpenAI | Unnamed OpenAI model described as more capable than GPT-6 Astra for the proof; GPT-6 Astra through Codex for the Lean formalization | Claimed | AI-led | mathematics |
| 2026-09-08 | Finite-time blowup for the unforced Euler equations from smooth compactly supported data | OpenAI | Unnamed OpenAI model described as more capable than GPT-6 Astra for the proof; GPT-6 Astra through Codex for the Lean formalization | Claimed | AI-led | mathematics |
| 2026-09-05 | Disproof of the Ibragimov-Iosifescu conjecture for phi-mixing sequences | OpenAI / Epoch AI | GPT-6 Astra (pre-release) | Claimed | AI-led | mathematics |
| 2026-09-03 | Lean disproof of Krempa's matrix form of the Koethe conjecture | OpenAI / Epoch AI | GPT-6 Astra (pre-release) | Formally verified | Autonomous | mathematics |
| 2026-09-03 | Erdos-Sos conjecture proved in Lean, in a form marginally weaker than the classical statement | OpenAI / Epoch AI | GPT-6 Astra (pre-release) | Formally verified | Autonomous | mathematics |
| 2026-09-03 | Counterexample to Smale's mean value conjecture at K = 1 | OpenAI / Epoch AI | GPT-6 Astra (pre-release) | Formally verified | Autonomous | mathematics |
| 2026-09-02 | Prime gaps at most 186, conditional on three unproved Lean axioms | OpenAI | GPT-6 Astra | Claimed | AI-led | mathematics |
| 2026-08-01 | Ten results in mathematics and theoretical computer science with Lean certificates | OpenAI | Astra | Formally verified | AI-led | mathematics |
| 2026-07-10 | Cycle double cover conjecture proved for all bridgeless multigraphs | OpenAI | GPT-5.6 Sol Ultra | Formally verified | AI-led | mathematics |
| 2026-05-20 | Disproof of the Erdős unit-distance conjecture | OpenAI | GPT-5 series reasoning model | Independently checked | AI-led | mathematics |
| 2026-01-13 | Erdős problem #728 resolved and formalized in Lean | OpenAI / Harmonic | GPT-5.2 Pro + Aristotle | Formally verified | Autonomous | mathematics |
| 2025-11-20 | Early science acceleration experiments with GPT-5 | OpenAI | GPT-5 | Peer reviewed | AI-assisted | computer-science |
| 2025-10-19 | GPT-5 "solved 10 Erdős problems": it located existing solutions | OpenAI | GPT-5 | Already known | Retrieval | mathematics |
| 2025-08-01 | Improved step-size bound in smooth convex optimization | OpenAI | GPT-5 Pro | Already known | AI-assisted | mathematics |