OpenAI Drops 722 Math Papers: Quasi-Riemann, BSD, Hilbert's 10th Solved

OpenAI has open-sourced 722 mathematical papers from an unpublished internal model, covering 17 fields with major breakthroughs including the quasi-Riemann hypothesis, full BSD formula, and Hilbert's tenth problem over rationals, averaging three hours of ChatGPT Pro compute per paper, while an independent advisory group of Fields medalists urges cautious verification.

PaperAgent
PaperAgent
PaperAgent
OpenAI Drops 722 Math Papers: Quasi-Riemann, BSD, Hilbert's 10th Solved

OpenAI recently released a massive collection of 722 mathematical manuscripts generated by an unpublished internal frontier model, all hosted in a single GitHub repository that garnered over 4,000 stars within hours. The papers are organized into 372 "families" — each a main result accompanied by supporting proofs, corollaries, or alternative arguments — spanning 17 branches of mathematics. On average, each paper consumed roughly three hours of ChatGPT Pro reasoning compute.

372 Families Across 17 Fields

The repository root contains a 41-page overview PDF mapping the entire mathematical landscape covered. A distribution chart shows the families spread across number theory, logic, probability, theoretical computer science, and other disciplines.

Headline Results

Quasi-Riemann Hypothesis: Proves zero-free regions for all Dirichlet L-functions (including the Riemann zeta function) in the half-plane Re(s) > 7/8, with an alternative proof reaching Re(s) > 11/12.

Full BSD Formula: Establishes the complete Birch–Swinnerton-Dyer leading-term formula for all elliptic curves with Selmer corank 0 or 1, including finiteness of the Tate–Shafarevich group.

Hilbert's Tenth Problem over ℚ: Negatively resolves the century-old question — no general algorithm exists to decide whether a polynomial with rational coefficients has a rational root.

Catalan's Constant Irrationality: Proves the irrationality of the famous Catalan constant.

Goldfeld Conjecture: Shows that in quadratic twist families of elliptic curves, analytic ranks 0 and 1 each occupy exactly half the density.

Random 3-SAT Threshold: Demonstrates the existence and computability of the satisfiability threshold for random 3-SAT formulas.

Critical Percolation No Infinite Clusters: Confirms the core conjecture that three-dimensional lattice percolation at criticality has no infinite clusters.

Exact Quantum Decomposition over Fixed Finite Gate Sets: Achieves exact synthesis of quantum circuits from a fixed finite gate set.

Toms–Winter Conjecture: Resolves a major classification problem in C*-algebra theory.

Independent Advisory Group (AGMAI)

The work was developed in consultation with the Advisory Group on Mathematics and Artificial Intelligence (AGMAI), an independent body formed under the Institute for Advanced Study. Its nine members include Fields medalists Timothy Gowers, Martin Hairer, and Edward Witten. Crucially, members accept no compensation from AI companies, and the group's mandate is open to all frontier AI labs.

On October 6, AGMAI issued a carefully worded statement:

Serving as advisors does not constitute endorsement of the results or the generation process.

Judgment of these works belongs solely to the mathematical community.

"Public release is only the first step — it is the beginning of human understanding and assimilation of this work, not its completion."

Mathematicians must retain the freedom to pose their own questions and explore their own directions; the future of mathematics cannot be merely "understanding AI-generated outputs."

Community Reaction

The release has ignited intense debate. Fields medalist Terence Tao, who previously delivered a public lecture "Mathematics in the Age of AI" at ICM 2026 and co-authored a declaration with 25 Fields medalists warning that "AI is destroying mathematics itself," is among the prominent voices urging rigorous verification. The mathematical community now faces the monumental task of checking hundreds of novel claims.

Resources

https://github.com/openai/math
https://x.com/OpenAI/status/2107596713791767021
https://agmai.org/
Original Source

Signed-in readers can open the original source through BestHub's protected redirect.

Sign in to view source
Republication Notice

This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactadmin@besthub.devand we will review it promptly.

OpenAImathematicsFields medalistsAGMAIAI-generated proofsBSD conjectureHilbert's tenth problemquasi-Riemann hypothesis
PaperAgent
Written by

PaperAgent

Daily updates, analyzing cutting-edge AI research papers

0 followers
Reader feedback

How this landed with the community

Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.