Companion page AI use in arXiv mathematics papers How many new arXiv mathematics papers show signs of AI writing, and how many say AI helped write them

arXiv · mathematics · January 2026 – 13 November 2026

In the wake of AI papers

Count new works on

In November 2026, through the 13th, 54 new arXiv mathematics papers follow up an AI-generated paper: 1.4% of the month’s papers, against 121 in October and 29 in September.

A paper follows up an AI-generated paper when its main result uses that paper’s results, extends them, adapts its method, or re-proves, explains or checks it. Papers that only cite one are not counted. See what counts.

Follow-up papers
236since the first, in June 2026
In November 2026
541.4% of the month’s papers, to 13 November; 45 of them follow up OpenAI’s Oct 6 release
AI-generated works followed up
124of the 938 AI-generated works on the list

Follow-up papers by month

New arXiv mathematics papers whose main result follows up an AI-generated paper, by the month of their first version. November 2026 runs to the 13th.

Follow up OpenAI’s Oct 6 releaseFollow up another AI-generated paperMonth in progress
Every month, with the share of the month’s papers
MonthPapersFollow up an AI-generated paperCite one, without following it upShare of the month’s papers
AllOpenAI’s Oct 6 releaseBy an author of the work
January 20263,51000000
February 20263,44000000
March 20264,32700010
April 20264,07600010
May 20264,28500010
June 20264,88030070.06%0.02–0.18%
July 20265,2091000110.19%0.10–0.35%
August 20265,8241902220.33%0.21–0.51%
September 20268,2462902610.35%0.24–0.50%
October 20268,7391218762201.38%1.16–1.65%
November 2026, to the 13th3,898544541051.39%1.06–1.80%

Papers counts every new preprint of the month in the 28 categories. The share is of that count, with a 95% interval. By an author of the work counts follow-up papers that share an author with every AI-generated work they follow up; they are included in the counts. The table starts in January 2026, the month of the earliest AI-generated paper on the list: no earlier paper can follow one up.

The most followed-up AI-generated works

Each AI-generated work with the number of papers that follow it up. Open a work to see those papers and what each one does with it. Each result family of OpenAI’s October 6 release counts as one work, and so does each of the ten results of OpenAI’s earlier “Ten advances”; ★ marks the release’s headline results. One paper studies the release as a whole and is listed under it.

  1. OpenAI’s release · family 003 ★The quasi-Riemann hypothesisOpenAI, 6 October 2026QRH Leaderboard: the best bounds for the quasi-Riemann hypothesis · community notes on oai-math-revised17follow-up papers

    The family’s manuscripts: The Quasi-Riemann Hypothesis: A Zero-Free Half-Plane ℜs > 7/8; The Quasi-Riemann Hypothesis (alternate 11/12 proof); Uniform exclusion of Landau–Siegel zeros.

    26 other papers cite it without following it up.

    • 2610.12234Slightly improved zero-free half-planes for the quasi-Riemann hypothesisBaiying Liu · math.NT · 8 October 2026Main result improves its 7/8 half-plane by varying its parameters
    • 2610.12803A conditional bound Λ ≤ 0.13963 for the de Bruijn–Newman constantCetin Hakimoglu-Brown · math.NT · 8 October 2026Main bound conditional on its 7/8 half-plane
    • 2610.15291Refining the 7/8 zero-free half-plane for Hecke and Dirichlet L-functionsHejian Sang · math.NT · 11 October 2026Main result improves its 7/8 half-plane, assuming its analytic inputs
    • 2610.15483On the OpenAI proof that Landau-Siegel zeros do not existWilliam D. Banks · math.NT · 11 October 2026Exposition of its Landau–Siegel argument with explicit constants
    • 2610.15559The Goldbach exceptional set: an exponent 7/12 and conditional refinementJános Pintz and Ingo Althöfer · math.NT · 11 October 2026Main theorem's zero-free half-plane hypothesis supplied by its uniform 7/8 theorem
    • 2610.18491Some applications of the quasi-Riemann hypothesis to class groups and small generators of number fieldsChristopher Frei, Efthymios Sofos and Martin Widmer · math.NT · 13 October 2026Main results conditional on the quasi-Riemann hypothesis, made unconditional by the release
    • 2610.20058Primes between 8th powers?Matt Visser · math.NT · 14 October 2026All main results conditional on its 7/8 half-plane
    • 2610.21099A simpler proof that there are no Landau–Siegel zerosBryce Kerr · math.NT · 15 October 2026Simplified proof of its Landau–Siegel theorem
    • 2610.21511Imaginary quadratic fields of class number at most 2000Iwao Kimura · math.NT · 15 October 2026Main result rests on making its Landau–Siegel argument explicit, with a Lean formalization
    • 2610.26756Fourth moments of prime exponential sums and exceptional sets for sums and differences of primesHuixi Li and Zhixin Liu · math.NT · 19 October 2026Main theorem conditional on a quasi-Riemann hypothesis with Θ < 11/12, supplied by its 7/8 half-plane
    • 2610.26835Zero-free half-planes for Hecke L-functions over number fieldsBaiying Liu · math.NT · 19 October 2026Extends its cubic-theta argument for the 7/8 half-plane to Hecke L-functions, using the primitive-roots manuscript's constructions
    • 2610.29470The exceptional set of Goldbach numbers within the set of squaresA. Perelli and J. Pintz · math.NT · 21 October 2026Main theorem's major-arc evaluation uses its 7/8 zero-free half-plane as an input
    • 2610.30476An explicit elementary proof of no Landau–Siegel zerosVictor Zhenyu Guo · math.NT · 21 October 2026Main theorem is an explicit elementary version following its Frobenius method for Landau–Siegel zeros
    • 2610.31402Landau-Siegel zeros over a base field and effective Brauer-SiegelWilliam D. Banks · math.NT · 22 October 2026Main theorems extend (Kerr's form of) its argument excluding Landau–Siegel zeros to a base field
    • 2610.36588Square-root logarithmic bounds for the arithmetic complexity of factorial multiplesLiang Zhang · math.NT · 27 October 2026Main theorem on the arithmetic complexity of multiples of k! conditional on a fixed zero-free half-plane
    • 2611.09386The exceptional set of Goldbach numbers among polynomial valuesI. Althöfer, D. Chen, A. Perelli and J. Pintz · math.NT · 6 November 2026Main theorem on 2φ(n) as a sum of two primes uses its 7/8 half-plane as the hypothesis of the fourth-moment estimates
    • 2611.16665Proof of the Quasi Riemann Hypothesis using ℚ[i]Xin Hang Ji · math.NT · 12 November 2026A shorter proof of a quasi-Riemann hypothesis for ζ (no zeros in Re s > 599/600), with quartic Gauss sums over ℤ[i] in place of its cubic ones
  2. Not on arXivA counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')Fable (Claude, Anthropic), posted by Levent Alpöge, 20 July 202613follow-up papers

    20 other papers cite it without following it up.

    • 2607.20210Graded Keller maps and the Jacobian ConjectureT. Shaska · math.AG · 22 July 2026Studies its grading, fibres and rational points, and shows that the sign pattern of a grading decides whether graded Keller maps can be counterexamples
    • 2608.23777An Explicit Counterexample to the Rank-Two Poisson ConjectureChristopher D. Long · math.RA · 22 July 2026A counterexample to the rank-two Poisson conjecture, built from its determinant-one normalization
    • 2607.20968An Explicit Characteristic-2 Counterexample to the Separable Jacobian ConjectureIrit Huq-Kuruvilla · math.AG · 23 July 2026A counterexample to the separable Jacobian conjecture in characteristic 2, by a modification of it (it does not itself specialize)
    • 2607.21572Generic degrees of real polynomial Keller maps with non-dense imagePiotr Migus · math.AG · 23 July 2026Every even generic degree at least 4 of real Keller maps with non-dense image, realized by a deformation family of it
    • 2607.22198A five-variable counterexample to the Hessian conjecture, and the low-dimensional status of the Jacobian and Hessian conjecturesGuowu Meng and Liang Yang · math.AC · 24 July 2026A five-variable counterexample to the Hessian conjecture, from its six-variable doubling by a one-variable Schur descent
    • 2608.00222Counterexamples to the Jacobian conjecture in dimensions greater than twoShuhong Gao · math.AG · 31 July 2026A self-contained account of its tangent-sweep mechanism, generalized to direction fields on hypersurfaces, giving counterexamples in dimensions above two
    • 2608.02863Orbits and Fields of Definition for Graded Keller MapsKyle Kistner and Tanush Shaska · math.AG · 3 August 2026Classifies graded Keller maps by their weights, with new families, and shows it is the unique graded Keller map in its degree box
    • 2608.05392The weak Markus–Yamabe conjecture fails in dimension 14Álvaro Castañeda, Gerardo Honorato and Francisco Valenzuela-Henríquez · math.AG · 5 August 2026The weak Markus–Yamabe conjecture fails in dimension 14, by a chain realization theorem applied to it
    • 2608.12543Collision-Generated Compression for Homogeneous Keller MapsThomas Prellberg · math.AG · 12 August 2026A 40-variable counterexample to Zhao's vanishing conjecture, from Thompson's 24-variable cubic-homogeneous reduction of it
    • 2608.27341Divisors in projective bundles over the projective line whose complement is affine spaceRemy van Dobben de Bruyn · math.AG · 27 August 2026A criterion for complements of divisors in projective bundles over ℙ¹ to be affine space, which explains its construction without coordinates
    • 2609.17795An elementary way to find a counterexample to the Jacobian ConjectureArno van den Essen · math.AG · 15 September 2026Finds it again by elementary means, up to linear changes of coordinates
    • 2610.24889Alpöge's counterexample to the JC, Economic Stabilization, and Zhao's Vanishing ConjectureMichiel de Bondt · math.AG · 17 October 2026Counterexamples to the homogeneous Jacobian conjecture in dimension 15 and to Zhao's vanishing conjecture, by economic stabilization of a tweaked version of it
    • 2611.03572Counterexamples to the Jacobian Conjecture from the viewpoint of lineabilityL. Bernal-González, M. C. Calderón-Moreno and J. A. Prado-Bassas · math.AG · 2 November 2026Studies the size of the family of counterexamples in dimension N ≥ 3, nonempty through it: lineability exactly 1, affine coneability, semigroups under composition
  3. Not on arXivThe (3,4,∞) modular family of 2-tori, completed at its three special points, is a complex structure on S⁶Claude and Levent Alpöge, August 20269follow-up papers

    3 other papers cite it without following it up.

    • 2608.29651Remarks on the Complex Structures on ℙ3 and S2×S4Ruiming Liang, Chenhan Liu and Yang Zhang · math.DG · 30 August 2026Assumes its complex structure on S⁶ and builds from it complex structures on ℙ³ and on S² × S⁴
    • 2609.11541A Complex Structure on A Rational Homology Six-Sphere with Two-TorsionZichang Wang · math.DG · 10 September 2026Adapts its construction, in Engel's form, to another elliptic surface, giving a simply connected rational homology six-sphere with H₂ ≅ H₃ ≅ ℤ/2
    • 2609.29666Compact Complex Threefolds Fibred by Two-Dimensional Complex ToriXiufan Yang · math.AG · 13 September 2026Carries out its (3,4,∞) construction for the monodromy types (4,6) and (3,6)
    • 2609.26706Conditional Oka Complex Structures of the Six-SphereZhangchi Chen · math.CV · 22 September 2026Proves that its threefolds are Oka manifolds
    • 2610.00250Oka Geometry and Deformations of Torus-Fibred Complex Six-SpheresYun-Heng Du, Bin Guo, Yuta Kusakabe and Song-Yan Xie · math.CV · 23 September 2026A locally miniversal family on its six-sphere, every member of which is Oka
    • 2609.33785A two-parameter family of complex structures on S6Jeff Viaclovsky · math.AG · 27 September 2026A two-parameter family of complex structures on S⁶ from another elliptic surface, not biholomorphic to its examples
    • 2609.38442Complex structures on S6Philip Engel · math.AG · 29 September 2026An exposition and a self-contained proof of its theorem
    • 2610.10451Combs and complex structures on the six-sphereJeff Viaclovsky · math.AG · 7 October 2026Modifies its structure to give complex structures on S⁶ of arbitrarily large local Kuranishi dimension
    • 2610.39102Logarithmic transformations on the six-sphereJeff Viaclovsky · math.AG · 29 October 2026Infinitely many choices of the multiplicities in its construction give complex structures on S⁶, pairwise biholomorphically inequivalent
  4. arXiv 2608.13637More than two thirds of the zeta zeros are simple and on the critical lineLevent Alpöge and Ralph Furman, August 20267follow-up papers

    8 other papers cite it without following it up.

    • 2609.02882A new proof that more than 2/3 of the zeros of the Riemann zeta function are simple and on the critical lineYouness Lamzouri · math.NT · 2 September 2026A new proof by a different method
    • 2609.33043More than 83.69% of the zeros of the Riemann zeta function are distinctKristian Muri Knausgård · math.NT · 27 September 2026Improves its proportion of distinct zeros, following its proof
    • 2610.08965More than 83.9% of the zeros of the Riemann zeta function are distinct and more than 67.35% are simple and on the critical lineKristian Muri Knausgård · math.NT · 6 October 2026Improves both of its proportions by the same energy method, with a Lean formalization
    • 2610.12853On the Imaginary Part of ζ′/ζFarzad Aryan · math.NT · 9 October 2026An exposition of its proof, beside a zero-side formula for the imaginary part of ζ′/ζ
    • 2610.22612Proof Perplexity: A Probabilistic Approach to Measuring non-triviality in Machine Generated ProofsMadhuparna Das and Soumya Banerjee · math.LO · 15 October 2026A measure of the non-triviality of formal proofs, applied only to its Lean formalization
    • 2610.22927Unconditional Bounds for Two, Three and Four Consecutive Moderate Gaps between Zeros of the Riemann Zeta FunctionCésar Angón, Erick Ross, Ze Sen Tang and Boyang Zhao · math.NT · 16 October 2026Adapts its method to consecutive moderate gaps between zeros
    • 2610.24612Human Exposition of Claude's 2/3 Theorem on the zeros of zetaJ. Arias de Reyna · math.NT · 17 October 2026A self-contained exposition of its proof, with a modification the author says the argument needs
  5. OpenAI · not on arXivFinite time blowup for Navier–StokesOpenAI, 8 September 2026the announcement · the Lean proofs7follow-up papers

    15 other papers cite it without following it up.

    • 2609.10085Finite-time divergence of kinetic-energy fluctuations in a Schrödinger-induction flowWeishuo Liu · math.AP · 9 September 2026Main theorem takes its forced blow-up solution as hypothesis and carries it into a constrained Schrödinger–induction model, where the kinetic-energy fluctuations diverge in finite time
    • 2609.10262Density of Forces Producing Navier–Stokes BlowupShaozhen Cao, Zhuoni Chi and Ping Nie · math.AP · 9 September 2026Inserts its blow-up solution into regular trajectories: smooth forces producing blow-up from rest are dense on the torus in time-integrated Hˢ exactly when s < 1/2
    • 2609.10269Singular Forces on the Whole Space: Sobolev Density Thresholds and Energy ApproximationShaozhen Cao and Zhuoni Chi · math.AP · 9 September 2026The counterpart on ℝ³: forces causing breakdown are dense in time-integrated Sobolev topologies exactly below a sharp threshold
    • 2609.20803Regularity of asymptotically axisymmetric solutions to the 3D Navier-Stokes equations with analytic forcingPeter Constantin, Mihaela Ignatova and Vlad Vicol · math.AP · 17 September 2026Takes two properties of its construction as hypotheses and proves regularity under a real-analytic force, so its force cannot be real analytic near the singular point (the paper is itself on the list)
    • 2609.35406Finite-Time Blowup for Navier-Stokes with Smooth Forcing, Part I: Construction of Self-Similar Solutions with Admissible Stress and Flat RemainderZhen Lei and Xiao Ren · math.AP · 28 September 2026An exposition of the profile construction of its manuscript
    • 2610.08144Navier-Stokes lost in translation: Why Lean verification of AI autoformalisation does not guarantee correct natural language proofsAlexander Bastounis, Fabian Circelli and Anders C. Hansen · math.AP · 6 October 2026Compares its natural-language proof with its Lean formalization and argues that they do not correspond, making no claim about the proof itself
    • 2611.06869Finite-Time Blowup for Navier-Stokes with Smooth Forcing, Part II: Residual Correction via Oscillatory PulsesZhen Lei and Xiao Ren · math.AP · 4 November 2026The second of two papers expounding its pulse corrections, claiming no original contribution
  6. OpenAI’s release · family 370The Lane–Emden and Hénon–Lane–Emden conjecturesOpenAI, 6 October 2026community notes on oai-math-revised6follow-up papers

    The manuscript: The Subcritical Hénon–Lane–Emden Conjecture.

    4 other papers cite it without following it up.

    • 2610.10188A proof of the fractional Lane-Emden conjectureLinfeng Mei and Juncheng Wei · math.AP · 7 October 2026Main theorem proved by adapting its Hénon–Lane–Emden method
    • 2610.22802The fractional Hénon-Lane-Emden conjecture and some critical nonexistence phenomenaKui Li, Haoyang Lu and Zhitao Zhang · math.AP · 16 October 2026Main theorem refines and extends its Hénon–Lane–Emden method to fractional orders
    • 2610.26656Weighted potential and Green systems with Dini coefficients and primitive growthHongtao Hu, Zhouyu Long and Wenming Zou · math.AP · 19 October 2026Proof of the main nonexistence theorem inherits and adapts its interval-pressure framework
    • 2610.27875Classification of positive solutions to the critical Lane-Emden systemYuxia Guo, Yichen Hu, Shaolong Peng and Tingfeng Yuan · math.AP · 20 October 2026Main classification proved by its source-layer interval comparison method
    • 2610.31167Liouville Theorems and Sharp Homogeneity Defects for Weighted Potential SystemsShengbing Deng and Bo Zhang · math.AP · 22 October 2026Main Liouville theorem proved within its source-layer interval-pressure framework
    • 2611.02448The polyharmonic Hénon–Lane–Emden conjecturePhuong Le · math.AP · 2 November 2026Extends its argument to integral systems with Riesz kernels, proving the polyharmonic Hénon–Lane–Emden conjecture
  7. arXiv 2607.25533Computable Quantification in Reflective Grounded ArithmeticBryan Ford, July 20265follow-up papers
    • 2608.16140Internalized Truth in Reflective Grounded Arithmetican author of the workBryan Ford · math.LO · 17 August 2026A sequel by the same author: an internal truth predicate for the system, in Isabelle/HOL
    • 2608.17862Idealizing Useful Fictions in Omega Grounded Arithmetican author of the workBryan Ford · math.LO · 18 August 2026A sequel by the same author: the system extended by an ω-grounded rule
    • 2610.34548Inductive Grounded Arithmetic: Goodstein, the Hydra, and Γ0an author of the workBryan Ford · math.LO · 25 October 2026A sequel by the same author: the system with an added induction rule, and its soundness, consistency and ordinal results
    • 2611.00176Coinduction in Grounded Arithmetican author of the workBryan Ford · math.LO · 30 October 2026A sequel by the same author: its system, through the inductive extension, with a coinduction rule and a dichotomy rule, with soundness and consistency
    • 2611.13697Gödel's Incompleteness Theorems in Grounded Arithmetican author of the workBryan Ford · math.LO · 10 November 2026A sequel by the same author: Gödel's incompleteness theorems for its system, resting on its soundness and consistency
  8. Not on arXivA Counting Proof for Erdős Problem 548GPT-6 Astra (OpenAI), run by Epoch AI, 3 September 2026the generated write-up (PDF) · the Lean proof5follow-up papers

    2 other papers cite it without following it up.

    • 2609.08012Kalai's Conjecture for Tight TreesDhruv Mubayi and Jacques Verstraete · math.CO · 7 September 2026Kalai's conjecture for tight trees in uniform hypergraphs, by extending its counting method (the proof itself found by GPT-6 Astra)
    • 2609.10987Erdős-Sós for digraphsDhruv Mubayi and Jacques Verstraete · math.CO · 10 September 2026The analogue for Eulerian digraphs, by its method (the proofs found by GPT-6 Astra; the paper is itself on the list)
    • 2609.15893A short proof of the Erdős–Sós ConjectureOliver Riordan and Alex Scott · math.CO · 14 September 2026A simplified version of its proof, with the extremal graphs and the antidirected-tree conjecture of Addario-Berry, Havet, Linhares Sales, Reed and Thomassé
    • 2609.21159Erdős-Sós via random cyclic orderingsBryce Frederickson · math.CO · 18 September 2026A simplified variant of its proof in the language of random cyclic orderings, with the same antidirected-tree conjecture
    • 2610.41286On the Turán number for the Cartesian product of an odd cycle and a treeYa-Hong Chen, Yong Lu, Quan Sun, Qi Wu and Long-Tu Yuan · math.CO · 30 October 2026The Turán number of the product of an odd cycle and a tree, with the Erdős–Sós theorem it proved as an input
  9. OpenAI’s release · family 011Prime-factor statistics of p-1OpenAI, 6 October 2026community notes on oai-math-revised5follow-up papers

    The family’s manuscripts: Weighted dilation graphs, smooth shifted primes and totient fibers; The Poisson-Dirichlet law for prime predecessors; Prime Predecessors with an Even Number of Prime Factors.

    • 2610.15674Coprime pairs with equal sum of divisors, assuming smooth shifted primesShisheng Li and Zhong Zhuang · math.NT · 11 October 2026Checks its argument for p+1; main conclusion conditional on it
    • 2610.20242On the Hardy–Littlewood–Chowla conjectureGuangshi Lü, Zhiwei Wang and Jie Wu · math.NT · 14 October 2026Type II estimate follows its determinant and memory method; root-counting lemma reproved from it
    • 2610.40439Shifted convolutions of divisor functionsMiguel N. Walsh · math.NT · 29 October 2026The asymptotic for shifted convolutions of divisor functions
    • 2611.12467An asymptotic formula for primes of the form u2+v2+1Yujiao Jiang, Yuk-Kam Lau and Guangshi Lü · math.NT · 9 November 2026Motohashi's conjecture on primes u² + v² + 1, with its harmonic-switching and high-moment method in the minor-arc bound
    • 2611.12756On the higher-order Titchmarsh divisor problemGuangwei Hu, Guangshi Lü, Zhiwei Wang and Jie Wu · math.NT · 9 November 2026Extends its Theorem 3.1 to every fixed nonzero shift, following its proof, for the asymptotic of Σ Λ(n)τₖ(n + h)
  10. OpenAI’s release · family 008The Deligne–Drinfeld conjectureOpenAI, 6 October 2026community notes on oai-math-revised4follow-up papers

    The manuscript: The Deligne-Drinfeld conjecture.

    2 other papers cite it without following it up.

    • 2610.15592The Deligne–Drinfeld conjecture after OpenAIThomas Willwacher · math.QA · 11 October 2026Simplified exposition of its proof
    • 2610.23719Confluence relations, Hoffman basis over ℤ(2), and the Deligne–Drinfeld conjectureMinoru Hirose · math.NT · 16 October 2026Abstract and title present the main theorem as yielding an alternative proof of its Deligne–Drinfeld theorem
    • 2610.41766An odd version of the Grothendieck–Teichmüller Lie algebra and the odd Deligne–Drinfeld conjectureFlorian Naef and Thomas Willwacher · math.QA · 30 October 2026An odd analogue of its Deligne–Drinfeld theorem
    • 2611.01745A remark on the AI proof of the Deligne–Drinfeld conjectureNikita Markarian · math.AG · 1 November 2026A geometric interpretation of the injective projection in its characteristic-two proof, making its reconstruction explicit
  11. OpenAI’s release · family 017 ★The irrationality exponent of π is 2OpenAI, 6 October 2026community notes on oai-math-revised4follow-up papers

    The manuscript: The irrationality exponent of pi is 2.

    1 other paper cites it without following it up.

    • 2610.10192Irrationality exponents of logarithms of positive rational numbersJingwen Liu, Kai Jiang and Pingwen Zhang · math.NT · 7 October 2026Main theorem proved by adapting the determinant method of its π paper
    • 2610.26539The irrationality exponent of the lemniscate period: A determinant proof with a Lean 4 formalizationGregory V. Chudnovsky · math.NT · 19 October 2026Main theorem proved by its π-exponent method transposed to the lemniscate, formalized on 828 modules of its Lean library
    • 2610.28079Irrationality exponents of logarithmic numbers: Values, quotients, and graded approximationGregory V. Chudnovsky · math.NT · 20 October 2026Main bounds obtained by adapting its interpolation and determinant method for the irrationality exponent of π
    • 2610.31112Irrationality Exponents of Real CM Elliptic Periods and Powers of the Lemniscate PeriodGregory V. Chudnovsky and David V. Chudnovsky · math.NT · 22 October 2026Main theorem proved by its separated-weight method for μ(π) = 2, partly formalized with its Lean library
  12. OpenAI’s release · family 079Local smoothing in three dimensionsOpenAI, 6 October 2026community notes on oai-math-revised4follow-up papers

    The manuscript: Critical local smoothing for the three-dimensional wave equation.

    1 other paper cites it without following it up.

    • 2610.12681The square function estimate for the cone in four dimensionsFaruk Temur · math.CA · 8 October 2026Main estimate derived from its local-smoothing and propagation statements
    • 2611.03214The Dirichlet problem for quaternionic Monge–Ampère equations on flat quaternionic quotientsXi-Nan Ma and Jingyong Zhu · math.AP · 2 November 2026Quaternionic Monge–Ampère Dirichlet problems on quotients of ℍ^m, of which the periodic theorem it uses for its local smoothing estimates is a special case
    • 2611.07085The square function estimates for cones in arbitrary dimensions and elliptic quadricsFaruk Temur · math.CA · 4 November 2026Reverse square function estimates for the cone and for spheres and other surfaces, from its local smoothing statements taken as hypotheses
    • 2611.11060The sharp Fourier extension estimate for the spherePang-hung Chung · math.AP · 8 November 2026The full strong-type Fourier extension estimate for the sphere in every dimension n ≥ 3, from two of its estimates
  13. OpenAI · Ten advances · result 3Non-sofic groupsOpenAI, 1 August 2026the announcement · the Lean proofs4follow-up papers

    14 other papers cite it without following it up.

    • 2608.02025A torsion-free non-sofic groupFrancesco Fournier-Facio · math.GR · 3 August 2026A finitely presented torsion-free non-sofic group, by applying its soficity criterion (its Proposition 2.3) to a small-cancellation construction
    • 2608.06222Nonsofic wreath products of residually finite groupsGabor Kun and Andreas Thom · math.GR · 6 August 2026Strengthens its Proposition 2.3 to show that generalized wreath products and doubles over infranormal property (T) subgroups are non-sofic
    • 2609.27795A conditional construction of a nonhyperlinear group and the centralizer problemAndreas Thom · math.GR · 14 September 2026A matrix counterpart of the Kun–Thom form of its mechanism, giving a nonhyperlinear group conditionally on the centralizer problem for matrix ultraproducts
    • 2610.08581Counterexamples to the continuous Aldous-Lyons conjectureTsachik Gelander · math.GR · 6 October 2026Main theorems take the existence of a finitely generated non-sofic group, credited to it with Fournier-Facio's and Kun–Thom's work, as input, giving non-co-sofic invariant random subgroups
  14. OpenAI · not on arXivPlanar Point Sets with Many Unit DistancesOpenAI, 20 May 2026the announcement4follow-up papers

    13 other papers cite it without following it up.

    • 2606.03419Optimizing Explicit Unit-Distance Lower-Bound CertificatesMichael T. M. Emmerich · math.OC · 2 June 2026Optimizes the parameters of its construction, in Sawin's explicit form, to certify the exponent 1.0152 for the number of unit distances
    • 2606.16738A counterexample to the near-quadratic Elekes–Rónyai expander conjecture over ℝJihao Liu · math.NT · 15 June 2026Disproves the near-quadratic Elekes–Rónyai expander conjecture over ℝ, with its tower of number fields (its Proposition 3.8) as the main input
    • 2606.27352Congruent copies of finite patterns in planar point setsShubhrajit Bhattacharya and Ritesh Goenka · math.CO · 25 June 2026Extends its lower bound from unit distances to congruent copies of any finite planar pattern
    • 2607.05374The Minkowski grid has robustly many repeated distancesSungchul Lee, Cosmin Pohoata and Daniel G. Zhu · math.CO · 6 July 2026A robust form of its counterexample by a construction of the same kind: every large subset of the grid has a much-repeated distance
  15. OpenAI · not on arXivA Proof of the Cycle Double Cover ConjectureOpenAI, July 20263follow-up papers

    9 other papers cite it without following it up.

    • 2607.15399OpenAI's proof of the Cycle Double Cover TheoremJim Geelen · math.CO · 16 July 2026A short exposition of its proof
    • 2607.16356A proof of the cycle double cover conjecture by OpenAI: An expositionSang-il Oum · math.CO · 17 July 2026A self-contained exposition of its proof, with modifications
    • 2607.24724Exponentially Many Circuit Double CoversRadek Hušek and Robert Šámal · math.CO · 27 July 2026Reformulates its proof and uses its system of linear equations to count circuit double covers
  16. OpenAI · not on arXivFinite time blowup for the Euler equationOpenAI, 8 September 2026the Lean proofs3follow-up papers

    10 other papers cite it without following it up.

    • 2610.02509Finite-time blowup of the Euler equation without forcingFan Zheng · math.AP · 1 October 2026An exposition of its construction of blow-up for the unforced Euler equations
    • 2610.06544Finite-time blowup of the IPM equation from smooth initial data without forcingFan Zheng · math.AP · 5 October 2026Adapts its construction to prove unforced blow-up for the incompressible porous media equation on the two-torus
    • 2610.06712Finite-time blowup for 2D unforced IPMMimi Dai · math.AP · 5 October 2026Unforced porous-media blow-up on the two-torus, taking from it the control of each stage's earlier history by the pressure Hessian
  17. OpenAI · not on arXivImproved Long Gaps Between PrimesOpenAI, 2 September 2026the Lean proof3follow-up papers

    2 other papers cite it without following it up.

    • 2609.38408Large gaps between Romanoff numbersArtyom Radomskii · math.NT · 29 September 2026Adapts its weight construction and its short-translates proposition to find ≫ log log X consecutive integers up to X that are not of the form p + 2ⁿ
    • 2610.10762Long intervals of consecutive composite values of polynomialsArtyom Radomskii · math.NT · 7 October 2026Adapts its short-translate construction to give c log X log log X consecutive composite values of an irreducible polynomial
    • 2610.14127A lower bound for additive complements to the primesArtyom Radomskii · math.NT · 9 October 2026B(X) ≫ log X log log X for every additive complement B of the primes, from the author's lemma adapting its construction
  18. OpenAI’s release · family 004 ★Hilbert’s tenth problem over ℚOpenAI, 6 October 2026community notes on oai-math-revised3follow-up papers

    The family’s manuscripts: Hilbert’s tenth problem over the rational numbers; A pointwise 2-converse for elliptic curves with rational two-torsion.

    1 other paper cites it without following it up.

    • 2610.2069186 unknowns for Hilbert's tenth problem over the rationalsZhi-Wei Sun · math.NT · 14 October 2026Main theorem refines its negative solution of Hilbert's tenth problem over Q to 86 unknowns
    • 2611.16130A Diophantine definition of ℤ in ℚAshay Burungale · math.NT · 12 November 2026Proves ℤ Diophantine in ℚ by an approach inspired by its manuscript, taking its rank-one curve over ℚ(√2) and adapting its sieve argument
    • 2611.17124Why should ℤ be Diophantine in ℚ?Alexander Kozachinskiy, Hector Pasten and Xavier Vidaux · math.NT · 12 November 2026A finitary version of its argument, conditional on finiteness of 2-primary Shafarevich–Tate groups, the abc conjecture and Dickson's conjecture
  19. OpenAI’s release · family 073The Falconer distance conjectureOpenAI, 6 October 2026community notes on oai-math-revised3follow-up papers

    The manuscript: The Falconer distance conjecture in all dimensions.

    2 other papers cite it without following it up.

    • 2610.22222A quantitative form of the planar Falconer distance theoremGuozheng Cheng, Xiang Fang, Menghan Li, Hongdou Qu and Chengbo Xiao · math.CA · 15 October 2026Main theorem develops its multiscale Falconer method into a quantitative form
    • 2611.13789On the proof of Falconer distance conjectureKotaro Inami, Jianbin Li, Yudong Liang, Bochen Liu and Francisco Romero Acosta · math.CA · 10 November 2026An exposition, for graduate readers, of a proof following its main idea
    • 2611.14867Falconer's conjecture through descent and chainingWill Burstein, Alex Iosevich and Alexia Yavicoli · math.CA · 11 November 2026A complete proof of its theorem, following its architecture with modifications, and a stronger conclusion on the Fourier dimension of the distance set
  20. OpenAI’s release · family 158The Euclidean plane cannot be colored with five colorsOpenAI, 6 October 2026community notes on oai-math-revised3follow-up papers

    The manuscript: The Euclidean plane is not five-colorable.

    3 other papers cite it without following it up.

    • 2610.15424Every n points in the plane contain (m1 + o(1)) n points with no unit distanceShisheng Li and Yong Su · math.CO · 11 October 2026Main theorem's only non-classical input is its Haar rigidity theorem
    • 2610.18907Ordinary and measurable colorings of Euclidean three-spaceMikhail Koshelev, Andrey Kupavskii and Vsevolod Voronov · math.CO · 11 October 2026Main theorem assumes its character-law rigidity theorem and adapts its transfer
    • 2611.14131Euclidean Ramsey Theory for Lines and Finite ConfigurationsRyan Hoff, Gary Hu and Yumo Liu · math.CO · 10 November 2026Its theorem on rotation-invariant measures from the five-colour proof is an input to all its Euclidean Ramsey results
  21. OpenAI’s release · family 248Thompson's group F is nonamenableOpenAI, 6 October 2026community notes on oai-math-revised3follow-up papers

    The manuscript: Thompson's group F is nonamenable.

    1 other paper cites it without following it up.

    • 2610.18360A note on the proof of nonamenability of FSang-hyun Kim and Yash Lodha · math.GR · 13 October 2026Exposition of its proof
    • 2610.27932Thompson's group F is not weakly amenableKoji Fujiwara · math.GR · 20 October 2026Main theorem deduced from its theorem that F is nonamenable
    • 2610.28473C*-selflessness and 𝒵-stability for Thompson's group FChris Bruce and Shirly Geffen · math.OA · 20 October 2026Both main theorems presented as applications of its nonamenability of F and its Corollary 1.3 on infinite tensor powers
  22. OpenAI’s release · family 337Sharp Cartan–Hadamard isoperimetry and rigidityOpenAI, 6 October 2026community notes on oai-math-revised3follow-up papers

    The family’s manuscripts: Generalized Cartan–Hadamard isoperimetry and Euclidean equality rigidity; Sharp integral fillings in CAT(0) spaces.

    • 2610.37817The generalized Cartan-Hadamard conjectureMarcos Agnoletto, Márcio Fabiano Da Silva, Stefano Nardulli and Reinaldo Resende · math.DG · 27 October 2026The generalized Cartan–Hadamard conjecture with rigidity, adapting its Michael–Simon inequality and balancing maps
    • 2611.01610Logarithmic Sobolev and exponential inequalities for submanifolds in Cartan–Hadamard manifoldsHongyu Chen and Li Ma · math.AP · 1 November 2026Extends its Michael–Simon type inequality to every codimension, for a logarithmic Sobolev inequality on submanifolds
    • 2611.13159Euclidean filling inequalities in CAT(0) spacesAlexander Lytchak · math.MG · 9 November 2026A streamlined version of its proof, extended to complete spaces, with rigidity in the equality case
  23. OpenAI · Ten advances · result 9Multicolor Ramsey numbersOpenAI, 1 August 2026the announcement · the Lean proofs3follow-up papers

    2 other papers cite it without following it up.

    • 2608.02522The Erdős-Hajnal conjecture for odd-girthRaphael Steiner · math.CO · 3 August 2026A single-exponential bound in the Erdős–Hajnal problem for odd girth, deduced from its superexponential lower bound for multicolour triangle Ramsey numbers
    • 2608.02537Multicolor Ramsey numbers of odd cycles are superexponentialRaphael Steiner · math.CO · 3 August 2026Modifies its construction to make the multicolour Ramsey numbers of every odd cycle superexponential
    • 2610.32789Triangle-Free Sets in Exact-Distance Hamming GraphsBence Csajbók, Padmini Mukkamala and Zoltán Lóránt Nagy · math.CO · 23 October 2026Its superexponential multicolour Ramsey lower bound as the input of the main ternary construction of triangle-free sets in Hamming graphs
  24. arXiv 2607.14379The small Davenport constant of the Heisenberg group of order 125Patrick White, July 20262follow-up papers
    • 2608.13747The small Davenport constant of the Heisenberg group of order 343Andreas Volkmann · math.CO · 13 August 2026The case p = 7 of the Godara–Sarkar conjecture by its method for p = 5
    • 2608.18651A Uniform Proof for the Small Davenport Constant of the Exponent-p Heisenberg GroupAndreas Volkmann · math.CO · 19 August 2026The Godara–Sarkar conjecture d(H_{p³}) = 3p − 3 for every odd prime p, generalizing its case p = 5
  25. Not on arXivθ(p_c) = 0 for Bernoulli bond percolation on ℤ^d in all dimensions d ≥ 2, in Lean 4 / MathlibClaude (Anthropic), directed by Justin Leder, 28 August 20262follow-up papers

    4 other papers cite it without following it up.

    • 2610.10502Notes on Kozma and Nitzan's proof of θ(pc) = 0Arthur Blanc-Renaudie · math.PR · 7 October 2026A shorter, human-written proof of the Kozma–Nitzan inequalities, which give θ(p_c) = 0, reconstructed from its proof
    • 2610.26419A connection inequality for the random-cluster modelJianping Jiang, Sike Lang and Yuxuan Sun · math.PR · 19 October 2026Extends the Kozma–Nitzan inequality, which it says AI proved first, to random-cluster measures with q ≥ 1, through Blanc-Renaudie's human version of the proof
  26. Not on arXivKozma–Nitzan Connection Inequalities for Bernoulli PercolationChatGPT 5.6 Sol and Claude Fable 5.1, prompted by Ahmed Bou-Rabee, 5 September 2026the verification page2follow-up papers

    1 other paper cites it without following it up.

    • 2610.10502Notes on Kozma and Nitzan's proof of θ(pc) = 0Arthur Blanc-Renaudie · math.PR · 7 October 2026A shorter, human-written proof of the inequalities it explains and extends, leaving to it the deduction of further Kozma–Nitzan conjectures
    • 2610.26419A connection inequality for the random-cluster modelJianping Jiang, Sike Lang and Yuxuan Sun · math.PR · 19 October 2026Extends the Kozma–Nitzan inequality, which it says AI proved first, to random-cluster measures with q ≥ 1, through Blanc-Renaudie's human version of the proof
  27. OpenAI’s release · family 002 ★The full BSD formula from low Selmer corankOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The family’s manuscripts: Exact Birch–Swinnerton-Dyer Formula from Low Selmer Corank; The Selmer converse for elliptic curves at every prime; The two-primary Birch–Swinnerton-Dyer formula in Selmer corank at most one.

    • 2610.09370On the algebraic and analytic ranks of the twin-prime elliptic curve y^2=x(x-2)(x-p)Ruihan Chen · math.NT · 7 October 2026Main theorem conditional on its low-corank BSD statement
    • 2610.31631Nonvanishing of 2-primary Kolyvagin classes via simultaneous quadratic twistsZhiqi Sun · math.NT · 22 October 2026Main index formula and nonvanishing theorem use its 2-converse and 2-primary BSD formula as inputs throughout
  28. OpenAI’s release · family 022The weak inhomogeneous Duffin–Schaeffer conjectureOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The manuscript: The Weak Inhomogeneous Duffin–Schaeffer Conjecture.

    • 2610.15809A note on the p-adic inhomogeneous Littlewood conjectureSam Chow and Owen Jones · math.NT · 11 October 2026Main theorem deduced from its weak inhomogeneous Duffin–Schaeffer theorem
    • 2611.01781Inhomogeneous Fiber Refinement of Gallagher's TheoremQian Li, Bo Tan and Qing-Long Zhou · math.NT · 1 November 2026Its theorem gives the divergence half of the main theorem on weighted inhomogeneous sets
  29. OpenAI’s release · family 039Nagata’s conjecture and maximal Seshadri constantsOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The family’s manuscripts: Nagata's conjecture for plane curves; Maximal Seshadri constants on arbitrary polarized surfaces; Maximal Multipoint Seshadri Constants in Higher Dimensions; Maximal multipoint Seshadri constants in positive characteristic.

    2 other papers cite it without following it up.

    • 2610.30050On OpenAI's proof of the Nagata conjectureJenia Tevelev · math.AG · 21 October 2026Expository notes on its proof of Nagata's conjecture
    • 2610.41931A Tropical Viewpoint on the Nagata–Biran–Szemberg ConjectureJuliette Bruce · math.AG · 30 October 2026Its Nagata–Biran–Szemberg theorem with an explicit threshold
  30. OpenAI’s release · family 062Projective contact classification and the LeBrun–Salamon conjectureOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The manuscript: Contact Fano manifolds and the LeBrun–Salamon conjecture.

    1 other paper cites it without following it up.

    • 2610.38736Transverse rational curves and the homogeneity of contact Fano manifoldsZhengyu Hu, Kefeng Liu and Xueyuan Wan · math.AG · 28 October 2026Homogeneity of every contact Fano manifold, which contains its classification, with a construction adapted from its manuscript
    • 2611.01423The classification of Fano contact manifolds revisitedBaohua Fu · math.AG · 1 November 2026An account, with simplifications, of its proof that every Fano contact manifold is an adjoint variety
  31. OpenAI’s release · family 072Brennan's conjecture and the integral-means spectrumOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The family’s manuscripts: Brennan's conjecture and sharp inverse-square integral means; A strict inverse-first-power bound for univalent functions.

    • 2610.23853Brennan's Conjecture and the Universal Integral Means SpectrumIlia Binder, Oleg Ivrii, Istvan Prause and Stanislav Smirnov · math.CV · 16 October 2026A shorter proof of a stronger form of its Brennan theorem, found while checking it
    • 2611.09697Geometric folding and maximal asymptotic varianceHaakan Hedenmalm · math.CV · 6 November 2026The maximal asymptotic variance bound for the schlicht class, by a finite-scale form of the folding method that grew out of its manuscripts
  32. OpenAI’s release · family 084The geometric case of the Erdős similarity conjectureOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The family’s manuscripts: The geometric case of the Erdős similarity conjecture; The dyadic case of the Erdős similarity conjecture.

    2 other papers cite it without following it up.

    • 2610.11449The Erdos similarity problem for null sequences with adjacent ratios bounded belowKan Jiang and Lifeng Xi · math.MG · 8 October 2026Main construction adapts its finite-tree construction
    • 2611.02454The geometric case of the Erdős Similarity Conjecture. A human exposition of the OpenAI proofAlex Iosevich and Alexia Yavicoli · math.PR · 2 November 2026A detailed exposition of its proof, related to earlier random-cell arguments and derandomized
  33. OpenAI’s release · family 090Triangular-lattice optimality, long-range Riesz and Coulomb energies, and spherical logarithmic energyOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The family’s manuscripts: An atomic certificate for triangular-lattice universal optimality; Universal optimality of the triangular lattice; A sharp Fourier certificate for planar circle packing; Triangular minimality for planar Coulomb renormalized energy.

    4 other papers cite it without following it up.

    • 2611.06278Uniqueness of the hexagonal lattice's universal optimalityMarkus Faulhuber and Philipp Petersen · math.MG · 4 November 2026Uniqueness: by its auxiliary functions, the triangular lattice is the only periodic configuration minimizing every Gaussian energy
    • 2611.17543The Triangular Abrikosov Constant and the Thomas–Fermi Γ-Limit of Rotating Bose–Einstein CondensatesGiovanni Di Fratta and Valeriy V. Slastikov · math.AP · 12 November 2026Its Gaussian-minorant theorem gives the lower bound identifying the interaction coefficient of rapidly rotating Bose–Einstein condensates
  34. OpenAI’s release · family 165The Harary–Hill and Zarankiewicz crossing-number formulasOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The family’s manuscripts: The crossing number of complete graphs; The crossing number of complete bipartite graphs.

    3 other papers cite it without following it up.

    • 2610.34444Proof of the Harary–Hill conjecture: an expositionMartin Balko · math.CO · 25 October 2026A simplified exposition of its proof of the Harary–Hill conjecture
    • 2611.03515The proof of the Harary–Hill conjecture: a detailed exposition and an extension to rotation systemsGelasio Salazar · math.CO · 2 November 2026A detailed exposition of its proof of the Harary–Hill conjecture for odd n, extending its polynomial method to rotation systems
  35. OpenAI’s release · family 170Sharp logarithmic exponents for off-diagonal Ramsey numbersOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The family’s manuscripts: The sharp logarithmic exponent of r(5,t); Sharp logarithmic exponents for fixed off-diagonal Ramsey numbers.

    • 2610.15293A note on the logarithmic exponent of Ramsey number r(4,t)Kaizhe Chen · math.CO · 11 October 2026Main result deduced from its Ramsey and independence bounds
    • 2610.17891The logarithmic exponent of the Ramsey number r(4,t)Qizhong Lin and Qiuyang Zhang · math.CO · 13 October 2026Main lower bound takes its r(5,t) bound as external input
  36. OpenAI’s release · family 196 ★A counterexample to Kaplansky’s zero-divisor conjectureOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The manuscript: A Torsion-Free Group Algebra with Zero Divisors.

    • 2610.17662Zero divisors in group algebras of torsion-free groups in odd characteristicS. V. Ivanov · math.GR · 13 October 2026Main theorem extends its counterexample to odd characteristic, using its Sections 3–5
    • 2610.22536Failure of direct finiteness for group algebras in every odd characteristicElena I. Bunina, Winston Lee, Alexei Miasnikov and Kyrylo Muliarchyk · math.GR · 15 October 2026Adapts the bounded-path and planar-localization arguments of its counterexample over 𝔽₂ to every odd prime p
  37. OpenAI’s release · family 197A torsion-free group algebra that is not directly finiteOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The family’s manuscripts: A Torsion-Free Group Algebra That Is Not Directly Finite; A Counterexample to Kaplansky's Direct-Finiteness Conjecture in Characteristic Two; A Counterexample to the Group-Ring Determinant Conjecture; A Counterexample to Kaplansky's Direct-Finiteness Conjecture in Odd Characteristic.

    • 2611.04599Kaplansky's conjectures fail for torsion-free hyperbolic groups in positive characteristicNidhal Mghirbi and Seth Douglas · math.GR · 3 November 2026Proves its torsion-free groups hyperbolic and adapts its construction to every odd characteristic, so that all four Kaplansky conjectures fail for hyperbolic groups
    • 2611.08268A Note on the OpenAI Construction of Counterexamples to the Kaplansky Direct-Finiteness ConjectureAndrei Jaikin-Zapirain · math.GR · 5 November 2026An exposition of its construction, adapted to every prime p: hyperbolic groups whose group algebra over 𝔽_p is not directly finite
  38. OpenAI’s release · family 213Critical percolation on every quasi-transitive graphOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The family’s manuscripts: Critical bond and site percolation on the cubic lattice; No percolation at criticality on quasi-transitive graphs.

    1 other paper cites it without following it up.

    • 2610.30063On the proof of the Kozma-Nitzan conjectureJean-Baptiste Gouéré · math.PR · 21 October 2026Exposition of its proof of the Kozma–Nitzan conjecture
    • 2611.01490No infinite cluster at criticality for long-range percolation with fast-decaying interactionsYan Ru Pei · math.PR · 1 November 2026Adapts its proof of no percolation at criticality on ℤ³ to long-range percolation on ℤᵈ
  39. OpenAI’s release · family 235Limiting random SAT thresholds, sharp variance and computabilityOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The family’s manuscripts: A Limiting Satisfiability Threshold for Every Fixed Clause Size; Linear Variance of the Random 3-SAT Hitting Time; Variance of the Random k-SAT Hitting Time; Computing the Random 3-SAT Threshold.

    • 2611.01338Threshold existence for random NAE-k-SAT and random hypergraph 2-colourabilityPavel Zakharov · math.CO · 31 October 2026Adapts its approach from random k-SAT to limiting thresholds for random NAE-k-SAT and hypergraph 2-colourability
    • 2611.10139The satisfiability conjecture for random k-SAT formulas. Human-friendly exposition of the recent OpenAI resolutionDavid Gamarnik · math.PR · 6 November 2026A reader-friendly exposition of its proof
  40. OpenAI’s release · family 296The generator problem for finite factorsOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The manuscript: Relative generation and the generator problem for finite factors.

    3 other papers cite it without following it up.

    • 2610.17411On the generators of separable unital C*-algebrasGeorge A. Elliott, Chun Guang Li and Zhuang Niu · math.OA · 12 October 2026Main results take its single-generation theorem as external input
    • 2610.28478A single generation theorem for simple unital C*-algebras with CPoU and comparisonXin Ma · math.OA · 20 October 2026Main theorem assumes its two-generation theorem for II₁ factors
  41. OpenAI’s release · family 369The hot spots conjecture for simply connected planar domainsOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The manuscript: Strict hot spots and absence of interior critical points on smooth simply connected planar domains.

    2 other papers cite it without following it up.

    • 2610.12622Removing Boundary Smoothness from the Hot-Spots Conclusion in the Simple-Eigenvalue CaseKrzysztof Burdzy · math.AP · 8 October 2026Main theorem is a reduction to its smooth-domain theorem, taken as input
    • 2610.32920Hot Spots: From Smooth to Nonsmooth Simply Connected Planar DomainsRodrigo Banñuelos · math.AP · 23 October 2026The hot-spots conclusion on simply connected domains of finite area, with its theorem for smooth domains as hypothesis
  42. OpenAI’s release · family 374Sharp one-third stability of Brenier mapsOpenAI, 6 October 2026community notes on oai-math-revised2follow-up papers

    The manuscript: Sharp One-Third Stability of Brenier Maps.

    3 other papers cite it without following it up.

    • 2610.39272Sharp potential and Brenier map W2-stability for log-concave sourcesYuanlong Ruan · math.AP · 29 October 2026Extends its one-third stability of Brenier maps to strongly log-concave sources
    • 2611.05534One-third Hölder stability of optimal transport maps from log-concave sourcesSongbo Wang and Xiaozhen Wang · math.AP · 3 November 2026Extends its one-third Hölder stability of Brenier maps to log-concave sources
  43. OpenAI · Ten advances · result 1High-dimensional sphere packingOpenAI, 1 August 2026the announcement · the Lean proofs2follow-up papers

    3 other papers cite it without following it up.

    • 2609.03736The half-rate linear programming bound for binary codes is 1/2−1/πAndrew Salmon · math.CO · 3 September 2026Proves the binary-code analogue of its sphere-packing theorem, the half-rate point 1/2 − 1/π of Delsarte's bound, following its mass-concentration argument
    • 2611.08224Fourier inequalities and sign uncertainty IIRoni Edwin · math.CA · 5 November 2026Adapts the argument of its Chapter 1 (the sign-uncertainty asymptotics) to an upper bound for the best constant in ‖f‖₂² ≤ w_d ‖f‖₁ ‖f̂‖₁ for radial functions
  44. OpenAI · Ten advances · result 2Binary and spherical codesOpenAI, 1 August 2026the announcement · the Lean proofs2follow-up papers

    2 other papers cite it without following it up.

    • 2609.03736The half-rate linear programming bound for binary codes is 1/2−1/πAndrew Salmon · math.CO · 3 September 2026Its upper bound adapts its spherical-code construction to binary codes, improving its value at the half-rate point to exactly 1/2 (the paper is itself on the list)
    • 2610.21263The Rate of the Delsarte Bound for Binary Codes and SpheresAndrew Salmon · math.CO · 15 October 2026Extends its constructions to find the exact rate of Delsarte's bound for binary and spherical codes
  45. arXiv 2604.06609Short proofs in combinatorics, probability and number theory IIBoris Alexeev, Moe Putterman, Mehtaab Sawhney, Mark Sellke and Gregory Valiant, April 20261follow-up paper

    2 other papers cite it without following it up.

    • 2607.15501Cycle lengths and chords under chromatic and degree constraintsXiaozheng Chen and Bo Ning · math.CO · 16 July 2026Extends its counterexample to Voss's conjecture from k = 4 to every k ≥ 5
  46. arXiv 2605.08122Undecidability problems for semifree DG algebrasCiprian Manolescu and Nick Rozenblyum, May 20261follow-up paper

    1 other paper cites it without following it up.

    • 2607.09667Undecidability of the quasi-isomorphism problem for semi-free dg algebras via the cobar constructionManuel Rivera · math.RA · 23 May 2026A new proof of its undecidability theorem by a different method
  47. arXiv 2606.05105Stochastically evolving ellipsoids with symmetriesElisha B. Abuya, Nihar Gargava and Yufei Zhao, June 20261follow-up paper

    2 other papers cite it without following it up.

    • 2610.12619A cyclotomic gluing construction for lattice sphere packingsan author of the workNihar Gargava, Matthias Gröbner and Michael Simkin · math.NT · 8 October 2026Extends its lower bound for the lattice sphere-packing density to every large dimension (a shared author)
  48. arXiv 2606.12776An AI Proof of 18-Variable Undecidability for Diophantine Equations over ℤ[i]Yuchen Ding and Junfeng Li, June 20261follow-up paper

    1 other paper cites it without following it up.

    • 2608.03992On Diophantine equations over the integer rings of quadratic fieldsZhi-Wei Sun · math.NT · 4 August 2026Hilbert's tenth problem over the integers of every quadratic field in 16 unknowns (15 for real fields), below its 18 over ℤ[i], using its Lemma 5
  49. arXiv 2607.04005K-polystable toric Fano varieties with small alpha invariantsJihao Liu and Ziwen Zhu, July 20261follow-up paper
    • 2608.14115The alpha spectrum of K-polystable toric ℚ-Fano varietiesXian Wu · math.AG · 14 August 2026Answers its question on α-invariants of toric Fano varieties, generalizing its construction
  50. arXiv 2607.10913τ-Hochschild (co)homology, the square of the Serre bimodule, and the Coxeter automorphism of the Tamarkin–Tsygan calculusMarco Armenta, July 20261follow-up paper
    • 2609.20220The Serre–Hochschild plane of a finite-dimensional algebraan author of the workMarco Armenta · math.KT · 26 July 2026Takes its Theorem 3.2 as a main input of several results (the same author)
  51. arXiv 2607.23302Optimality of Wouter van Doorn's Upper Bound for the Mayer-Erdős Farey ProblemRicky Cipollini, July 20261follow-up paper
    • 2608.15681An exact formula for Erdős' problem 1005Yanmohan Wang, Mingxu Xie and Ziyuan Zhao · math.NT · 16 August 2026Makes its asymptotic for Erdős problem 1005 exact, following its framework
  52. arXiv 2607.24483The exact solution of Bellman's lost-in-a-forest problem for the golden gnomonAlexander Temerev and Alessio Doria, July 20261follow-up paper

    3 other papers cite it without following it up.

    • 2609.04591Balanced Support Calibrations for Moser's Worm Problem: Exact Certificate and Bound of Triangular CoverZhipeng Deng · math.MG · 4 September 2026A lower bound in Bellman's escape problem by its balanced-support method and its Appendix A
  53. arXiv 2608.08752Overlap distribution of the critical Sherrington-Kirkpatrick modelHang Du and Brice Huang, August 20261follow-up paper

    2 other papers cite it without following it up.

    • 2609.14273Disorder transfer for the critical SK overlap lawYan Ru Pei · math.PR · 13 September 2026Transfers its Gaussian overlap limit to other disorder, assuming it
  54. arXiv 2608.16140Internalized Truth in Reflective Grounded ArithmeticBryan Ford, August 20261follow-up paper

    2 other papers cite it without following it up.

    • 2611.13697Gödel's Incompleteness Theorems in Grounded Arithmetican author of the workBryan Ford · math.LO · 10 November 2026A sequel by the same author: revises and extends its Gödel material, on its codes of terms and proofs and its proof checker
  55. arXiv 2608.17862Idealizing Useful Fictions in Omega Grounded ArithmeticBryan Ford, August 20261follow-up paper

    2 other papers cite it without following it up.

    • 2611.00176Coinduction in Grounded Arithmetican author of the workBryan Ford · math.LO · 30 October 2026A sequel by the same author: one of the new systems contains it, and a main theorem derives its omniscience rule from the dichotomy rule
  56. arXiv 2608.23132The Cartan determinant conjecture for representation-finite algebrasHaruhisa Enomoto, August 20261follow-up paper
    • 2610.14638The Cartan determinant conjecture for representation-finite algebras over perfect fieldsLiang Chen · math.RT · 10 October 2026Extends its theorem (Cartan determinant one for representation-finite algebras of finite global dimension) from algebraically closed to perfect fields, taking its grading over the algebraic closure as input
  57. arXiv 2609.01753The entropy production is not always monotone for hard spheres or for Maxwell moleculesLuis Silvestre, September 20261follow-up paper

    1 other paper cites it without following it up.

    • 2609.06290The entropy production is not always monotone for the Kac equationWeinan Wang · math.AP · 5 September 2026Adapts its two-temperature construction to the Kac equation
  58. arXiv 2609.03736The half-rate linear programming bound for binary codes is 1/2−1/πAndrew Salmon, September 20261follow-up paper
    • 2610.21263The Rate of the Delsarte Bound for Binary Codes and Spheresan author of the workAndrew Salmon · math.CO · 15 October 2026Extends its rate of Delsarte's bound to every distance and to spherical codes (the same author)
  59. arXiv 2609.05991A sub-Riemannian endpoint map with no regular valuesAntonio Lerario, Luca Rizzi and Daniele Tiberio, September 20261follow-up paper
    • 2610.24080An equiregular counterexample to the Sard conjecture for sub-Riemannian endpoint mapsBangjie Wang and Igor Zelenko · math.DG · 16 October 2026Answers the question its counterexample to the sub-Riemannian Sard conjecture left open
  60. arXiv 2609.10487Nonmaximal sums of maximally monotone operators under Rockafellar's constraint qualificationWeifeng Yang, September 20261follow-up paper
    • 2609.13906Failure of Rockafellar's Sum Conjecture via an Explicit Non-(FPV) OperatorRadu Ioan Bot · math.FA · 12 September 2026Simplifies and explains its counterexample to Rockafellar's sum conjecture
  61. arXiv 2609.12427Polynomial growth of Bohnenblust–Hille constants on the Hamming cubePaata Ivanisvili, September 20261follow-up paper

    1 other paper cites it without following it up.

    • 2610.15363Polynomial Bohnenblust–Hille Bounds and Hardy–Littlewood Inequalities for Finite Local SystemsAnderson Barbosa, Daniel Núñez-Alarcón, Daniel Pellegrino and Anselmo Raposo · math.FA · 11 October 2026Improves the polynomial bound on the Boolean Bohnenblust–Hille constants that it first proved to C·d^(5/2)(1 + log d)³, adapting its contraction argument
  62. arXiv 2609.16470Extending the Córdoba-Martínez-Zoroa IPM Blow-Up to Uniformly Space-Time Smooth ForcingLevent Alpöge, Tristan Buckmaster and Matei P. Coiculescu, September 20261follow-up paper

    5 other papers cite it without following it up.

    • 2610.06544Finite-time blowup of the IPM equation from smooth initial data without forcingFan Zheng · math.AP · 5 October 2026Removes the forcing from its blow-up for the incompressible porous media equation
  63. arXiv 2609.17760Every graph with no K7 =  minor is 6-colorableZdeněk Dvořák, Sergey Norin and Neil Rahman, September 20261follow-up paper
    • 2609.26041A sharp density bound for 5-connected graphs with no K7 =  minorCaibing Chang, Zijian Deng, Qinfei Tang and Caihong Yang · math.CO · 22 September 2026Proves its Conjecture 1.4 by adapting its reduction method
  64. arXiv 2609.18266Hidden Convexity via Symmetric Displacement Covers: Mixed-Integer Quadratic Programming and Integer Cubic Programming in Fixed DimensionCinar Ari and Robert Hildebrand, September 20261follow-up paper

    2 other papers cite it without following it up.

    • 2611.14837Two-Stage Mixed-Integer Quadratic Programming in Polynomial TimeAlberto Del Pia · math.OC · 10 November 2026Two-stage mixed-integer quadratic programs in polynomial time, generalizing its algorithm for mixed-integer quadratic programming in fixed dimension and using it for the subproblems
  65. arXiv 2609.19214A Strongly Aperiodic Monotile in Three DimensionsIoannis Tsiokos, September 20261follow-up paper
    • 2610.00916Towards Strongly Aperiodic Monotiles in Higher DimensionsDmitry Kamenetsky · math.MG · 1 October 2026Extends its three-dimensional strongly aperiodic monotile to ℝᴺ
  66. arXiv 2609.20803Regularity of asymptotically axisymmetric solutions to the 3D Navier-Stokes equations with analytic forcingPeter Constantin, Mihaela Ignatova and Vlad Vicol, September 20261follow-up paper

    2 other papers cite it without following it up.

    • 2610.25531Regularity of asymptotically axisymmetric Navier-Stokes flows with rotationally invariant curl of the forceBenjamin Ingimarson, Igor Kukavica and Qi Xu · math.AP · 18 October 2026Answers the question of its Remark 2.5
  67. arXiv 2609.26732Solution of the Mumford-Shah conjectureFrancesco Deangelis, September 20261follow-up paper

    1 other paper cites it without following it up.

    • 2610.33900A remark on the global minimality of the cracktipan author of the workFrancesco Deangelis, Camillo De Lellis and Matteo Focardi · math.AP · 24 October 2026A short proof that the cracktip is a global minimizer, with its resolution of the Mumford–Shah conjecture as the interior regularity input (shares an author with it)
  68. arXiv 2609.28238Second-Order Stationarity with Common Random Losses: Matching Tolerance BoundsWendao Wu, Haihan Zhang, Chenheng Zhang, Yanyi Li, Chunyuan Zheng, Cong Fang, Haoxuan Li and Zhouchen Lin, September 20261follow-up paper
    • 2610.16832Matching Complexity for Second-Order Stationary Point with Sample Reusean author of the workQian Chen, Wendao Wu, Haihan Zhang, Chenheng Zhang, Chunyuan Zheng, Cong Fang, Haoxuan Li and Zhouchen Lin · math.OC · 12 October 2026Extends its techniques to reusable samples (shared authors)
  69. arXiv 2609.29262The Marcus-Minc Transform InequalityYair Lavi, September 20261follow-up paper
    • 2609.31794The Wang Transform Inequalityan author of the workYair Lavi · math.CO · 25 September 2026Takes its lower estimate and bounds as inputs (the same author)
  70. arXiv 2609.29449Linear crossing families via dual levelsTim Gehrunger, September 20261follow-up paper
    • 2610.38658Larger Crossing FamiliesJoachim Orthaber and Patrick Schnider · math.CO · 28 October 2026Improves its bound for crossing families in line-separated sets, giving crossing families of size about n/79
  71. arXiv 2610.00245Matching Lower Bounds for Randomized First-Order Methods in Hessian-Lipschitz Nonconvex OptimizationHaihan Zhang, Wendao Wu, Chenheng Zhang, Yanyi Li, Chunyuan Zheng, Cong Fang, Haoxuan Li and Zhouchen Lin, October 20261follow-up paper

    1 other paper cites it without following it up.

    • 2611.05807Near-Optimal Complexity of Finding Approximate Local Minima of Nonconvex Finite SumsQihao Zhou · math.OC · 4 November 2026Replicates its spectral hard instance across components for the main theorem's lower bound against unrestricted randomized algorithms
  72. arXiv 2610.12434Global asymptotic stability of homogeneous polynomial vector fields is undecidableIvan O. Shevchenko, Jun Liu and Xinzhi Liu, October 20261follow-up paper
    • 2610.26832Controllability of polynomial systems is undecidable and not recursively enumerablean author of the workJun Liu and Ivan O. Shevchenko · math.OC · 19 October 2026Takes its Σ⁰₁-completeness of global asymptotic stability as the input of its main theorem, that global controllability of polynomial control systems is undecidable and not recursively enumerable (two of the same authors)
  73. arXiv 2610.25672On The Manin ConstantZhuoni Chi, Shurui Liu and Ping Nie, October 20261follow-up paper
    • 2610.31631Nonvanishing of 2-primary Kolyvagin classes via simultaneous quadratic twistsZhiqi Sun · math.NT · 22 October 2026Uses its theorem that the Manin constant of an optimal elliptic curve over ℚ is ±1 to normalize the modular parametrization in its main nonvanishing criterion for 2-primary Kolyvagin classes
  74. arXiv 2610.31052There is no universal Fσ equivalence relationRuijun Wang, October 20261follow-up paper
    • 2611.02531No Borel upper bound for the Fσ equivalence relationsan author of the workRuijun Wang · math.LO · 2 November 2026A sequel by the same author: no Borel equivalence relation of any complexity bounds all F_σ equivalence relations, strengthening its theorem by another proof
  75. arXiv 2610.34122The sharp leading asymptotics of the self-energy of the spinless Pauli-Fierz modelShinichiro Futakuchi, October 20261follow-up paper
    • 2610.34125The effective mass of the Pauli-Fierz electron: a cutoff-independent lower bound and the breakdown of mass renormalizationan author of the workShinichiro Futakuchi · math-ph · 25 October 2026Takes its sharp self-energy asymptotics as the input of one of its two main theorems, a lower bound of order λ^(8/7−ε) on the growth of the effective-mass ratio (the same author)
  76. arXiv 2610.39269Linearize or Discretize First? Error Estimates under Mesh-Uniform Regularity and Exponential Stability for the KdV Equation with Boundary DissipationKangsheng Liu, October 20261follow-up paper
    • 2611.12585A Large Gap between Galerkin Errors with and without Mesh-Uniform Regularity for the KdV Equation with Boundary Dissipationan author of the workKangsheng Liu · math.NA · 9 November 2026A sequel by the same author: shows that its error bounds for the weighted Hermite–Galerkin scheme fail for the standard scheme
  77. Not on arXivSmooth counterexamples to the Loewner and Carathéodory conjectures (X post)Levent Alpöge, John-Paul Smith and Claude (Anthropic), 19 August 20261follow-up paper
    • 2610.34036A simple construction of smooth counterexamples to the Loewner and Carathéodory conjecturesNaoya Ando and Masaaki Umehara · math.DG · 25 October 2026A simpler construction of its smooth counterexamples, by a variant of it with a trigonometric seed
  78. Not on arXivDisproof of the Köthe conjecture (Krempa's matrix form)GPT-6 Astra (OpenAI), run by Epoch AI, 3 September 2026the write-up (arXiv 2609.07996)1follow-up paper
    • 2609.15080The Köthe conjecture via point modulesB. Greenfeld, B. King and L. Vendramin · math.RA · 14 September 2026A new, short counterexample to the Köthe conjecture by another route, the point-module approach, after recalling that its counterexample came first
  79. Not on arXivA Tilted Residue-Class Construction for Long Prime-Free IntervalsGPT 5.6 Sol (OpenAI), prompted by DottedCalculator, 25 August 2026the proof claim and its discussion · Erdős problem #4 · B. Alexeev's Lean formalization1follow-up paper

    1 other paper cites it without following it up.

    • 2609.28253A tyro's approach to the tilted sieve: beyond the Erdős–Rankin boundTristan Freiberg · math.NT · 23 September 2026An elementary exposition of its tilted sieve and of its covering of composite survivors, without Maynard sieve weights or hypergraph covering, giving one residue class modulo each prime p ≤ x that covers an interval of length ≫ x log x/((log₂ x) log₃ x) (the paper is itself on the list)
  80. OpenAI’s releaseThe release as a whole, with its Lean libraryOpenAI, 6 October 20261follow-up paper

    8 other papers cite it without following it up.

    • 2610.25357Studying proof data with proof theory: the case of AI-generated mathematicsSafal Raman Aryal · math.LO · 18 October 2026Main measurements are made on 96 of its Lean projects as the principal corpus of AI-written proofs (the release studied as data, not used mathematically)
  81. OpenAI’s release · family 006Goldfeld’s conjecture: densities and mean analytic rankOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Goldfeld's analytic density conjecture and the 2-converse for elliptic curves; The mean analytic rank of quadratic twists of elliptic curves.

    • 2611.11566A rank-zero 2-converse to the Gross–Zagier and Kolyvagin theorem for elliptic curves with abelian two-division fieldAshay Burungale and Ye Tian · math.NT · 8 November 2026A self-contained proof, by its quadratic twist deformation method, that a finite 2^∞-Selmer group gives L(E,1) ≠ 0 when the two-division field is abelian
  82. OpenAI’s release · family 007Ordinary two-point correlations and the corrected Elliott conjectureOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: Ordinary two-point correlations of multiplicative functions.

    1 other paper cites it without following it up.

    • 2610.16394Dynamical generalizations of the two-point Chowla conjectureBiao Wang and Shaoyun Yi · math.NT · 12 October 2026Main theorem generalizes its two-point Chowla theorem and uses its estimates in the proof
  83. OpenAI’s release · family 025Short Egyptian fractionsOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: Short Egyptian fractions.

    • 2611.08819Reflexive polytopes can have doubly exponentially many verticesJohannes Hofscheier and Benjamin Nill · math.CO · 6 November 2026Its short Egyptian fractions, in Bloom and van Doorn's form, yield reflexive polytopes with doubly exponentially many facets
  84. OpenAI’s release · family 029Primitive roots for every admissible integer baseOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Primitive roots for every admissible integer base; Simultaneous primitive roots: a conditional lower bound for prime bases.

    2 other papers cite it without following it up.

    • 2610.24900All totally real abelian number fields of class number one are EuclideanHarald Grobner · math.NT · 17 October 2026Main theorem takes its zero-free region and its Proposition 2.1 as black-box inputs
  85. OpenAI’s release · family 032 ★Hodge and Kuga–Satake results for all projective K3 surfacesOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper
  86. OpenAI’s release · family 034Log abundance for compact Kähler spaces under logarithmic Iitaka subadditivityOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper
  87. OpenAI’s release · family 041Hyperkähler SYZ and projective-space basesOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Projective-space bases of Lagrangian fibrations; The strong hyperkähler SYZ conjecture.

    3 other papers cite it without following it up.

    • 2610.20496Smoothness criterion for Fano orbifolds and bases of Lagrangian fibrationsZhiyu Liu · math.AG · 14 October 2026Main theorem is a simpler proof of the second part of its result, combining its ideas with the author's
  88. OpenAI’s release · family 042Oka classification for minimal compact complex surfaces: Kodaira dimension zero and class VIIOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: Every complex K3 surface is Oka.

    1 other paper cites it without following it up.

    • 2610.10475A Strong Dominability Criterion and an Oka Union TheoremYun-Heng Du, Bin Guo, Peng-Chao Wang and Song-Yan Xie · math.CV · 7 October 2026Main theorem extends its Theorem 3.1 to all dimensions using its constructions
  89. OpenAI’s release · family 054Irrational cubic fourfolds with Hodge-theoretic and categorical K3 associationsOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: Irrational cubic fourfolds with geometric K3 categories.

    1 other paper cites it without following it up.

    • 2610.40811Integral Hodge structures and atoms of fourfoldsJérémy Guéré · math.AG · 29 October 2026A new proof of a theorem of its manuscript on irrational cubic fourfolds
  90. OpenAI’s release · family 057Fundamental groups of special complex varieties and root orbifoldsOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: The abelianity conjecture for special compact Kähler manifolds; Two-step monodromy of special quasi-projective varieties; A conditional abelianity theorem for special fourfold pairs with a half-weight divisor.

    2 other papers cite it without following it up.

    • 2611.14283Virtual Two-Step Nilpotency of the Fundamental Groups of Special Quasi-Projective VarietiesTianshu Jiang and Jiayu Li · math.AG · 10 November 2026Adapts its proof of Campana's abelianity conjecture to special quasi-projective varieties: fundamental group virtually nilpotent of class at most two
  91. OpenAI’s release · family 063The generalized Mukai conjectureOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: The generalized Mukai conjecture.

    2 other papers cite it without following it up.

    • 2611.05452On Fujita's Advanced Generalized Mukai ConjectureDuo Li · math.AG · 3 November 2026Fujita's advanced generalized Mukai conjecture, from its generalized Mukai inequality and its Gromov–Witten correspondences
  92. OpenAI’s release · family 076Real ultraflat Littlewood polynomials and unbounded binary merit factorsOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Ultraflat real Littlewood polynomials; Nearly minimal maxima and positive minima of Littlewood polynomials; Asymptotically minimal maxima of real Littlewood polynomials.

    • 2610.30797OpenAI partially disproves Littlewood's 1961 Lα-flat theoremel Houcein el Abdalaoui · math.NT · 22 October 2026Presents and analyses its counterexample to Littlewood's theorem
  93. OpenAI’s release · family 082Annular variation and dyadic absolute bounds for the triangular Hilbert transformOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Annular variation of the triangular Hilbert transform at the symmetric point; An L³ bound for the dyadic triangular Hilbert form; The maximal triangular Hilbert transform at the symmetric point.

    1 other paper cites it without following it up.

    • 2610.26388Pointwise convergence of bilinear ergodic averages for commuting transformationsZhuowen Guo, Rongzhong Xiao and Shuhao Zhang · math.DS · 19 October 2026Proof of the main theorem adapts its matrix-energy and Gaussian heat-flow method
  94. OpenAI’s release · family 085Endpoint Sobolev regularity of centered disk averagesOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: An Endpoint Gradient Bound for the Centered Disk Maximal Operator.

    • 2610.35760Endpoint gradient bounds for centered convex maximal operators and probability kernelsZhouyu Long · math.CA · 26 October 2026Extends its planar endpoint gradient bound for the centered disk maximal operator to convex bodies in every dimension, adapting its method
  95. OpenAI’s release · family 086An L3 bound for the trilinear Hilbert transformOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: An <i>L</i><sup>3</sup> bound for the trilinear Hilbert transform.

    2 other papers cite it without following it up.

    • 2610.39096Sparse domination and an extended Lp range for the trilinear Hilbert transformBingyang Hu · math.CA · 29 October 2026Sparse bounds for the trilinear Hilbert transform from its local estimate and stopping-time argument, extending its L³ × L³ × L³ → L¹ bound
  96. OpenAI’s release · family 087 ★The Mahler conjectures, functional inequalities and polar-product symplectic widthOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: The symmetric Mahler conjecture and its equality cases; The Mahler Conjecture for General Convex Bodies; Symplectic Balls in Symmetric Polar Products.

    1 other paper cites it without following it up.

    • 2610.34396A note on consequences of Mahler's inequalitiesYanlu Lian and Fei Xue · math.MG · 25 October 2026Makai's inequality in every dimension, derived from its symmetric Mahler theorem
  97. OpenAI’s release · family 091Logarithmic and Lp Brunn–Minkowski inequalities and the B-conjectureOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: The logarithmic Brunn–Minkowski conjecture.

    3 other papers cite it without following it up.

    • 2610.10152The centered logarithmic Brunn-Minkowski inequalityYude Liu, Qiang Sun and Ge Xiong · math.MG · 7 October 2026Main theorem extends its symmetric logarithmic Brunn–Minkowski proof to centered bodies, starting from its argument
  98. OpenAI’s release · family 146Positive metric entropy for the standard mapOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: Positive Metric Entropy for the Standard Map at Large Parameters.

    • 2611.09361Positivity of metric entropy of the standard map at large parametersS. Slijepčević · math.DS · 6 November 2026An explicit entropy bound of order log k for the standard map, following its proof's architecture
  99. OpenAI’s release · family 156Borsuk's conjecture fails in dimension nineOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: A nine-dimensional counterexample to Borsuk's covering assertion.

    4 other papers cite it without following it up.

    • 2611.08198A diameter-one set in ℝ5 that cannot be covered by six diameter-one ballsA. Arman, A. Bondarenko, A. Prymak and D. Radchenko · math.MG · 5 November 2026Its nine-dimensional Borsuk construction carried to ℝ⁵: a set of diameter 1 not covered by six balls of diameter 1
  100. OpenAI’s release · family 180Barnette’s Hamiltonian-cycle conjectureOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: Paired states and Hamiltonian cycles in cubic bipartite planar graphs.

    1 other paper cites it without following it up.

    • 2610.30490A note on the proof of Barnette's conjectureZhentao Li · math.CO · 21 October 2026Exposition of its proof of Barnette's conjecture
  101. OpenAI’s release · family 184Correspondence coloring with a fixed forbidden subgraphOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Correspondence coloring graphs with a forbidden clique; A logarithmic independence bound for clique-free graphs.

    • 2611.04297Fractional colorings of Kr-free graphs with local demandsEwan Davies, Abhishek Dhawan, Abhishek Methuku, Huy Nguyen and Jaehyeon Seo · math.CO · 3 November 2026The Kelly–Postle conjecture on fractional colouring with local demands, by adapting its triangle-weight argument
  102. OpenAI’s release · family 195A counterexample to the small Cohen–Macaulay module conjectureOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: A Complete Local Domain without a Small Cohen–Macaulay Module.

    • 2610.30027The recent counterexamples to the Eisenbud-Schreyer-Weyman Ulrich problem (2003) and to Hochster's small Cohen-Macaulay conjecture (1973): a survey, with a generalization to arbitrary dimensionCristian Anghel · math.AG · 21 October 2026Survey expounding its small Cohen–Macaulay counterexample, whose new results generalize its argument to every dimension
  103. OpenAI’s release · family 199Counterexamples to Auslander–Reiten, Tachikawa and related homological conjecturesOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: An explicit counterexample to the Auslander-Reiten conjecture; A counterexample to Tachikawa's second conjecture.

    1 other paper cites it without following it up.

    • 2611.12176From Frobenius seeds to self-orthogonal modules over symmetric algebrasHongxing Chen · math.RT · 9 November 2026Generalizes its counterexample to Tachikawa's second conjecture to a construction from any Frobenius algebra with a suitable module
  104. OpenAI’s release · family 215Canonical O(3) continuum limit and exact O(4) mass asymptoticsOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: The canonical massive continuum limit of the two-dimensional O(3) model; An Isolated Particle Pole for the Two-Dimensional O(3) Spin Field; Exact mass asymptotics for the two-dimensional O(4) lattice model; Sharp mass bounds for the two-dimensional O(4) model; Exponential decay in two-dimensional classical O(n) models.

    2 other papers cite it without following it up.

    • 2611.01532The massive phase of the two-dimensional Heisenberg model is stable under small anisotropyYan Ru Pei · math-ph · 1 November 2026Adapts its proof, with its annulus estimate as input, to show that exponential decay survives small anisotropy of the couplings
  105. OpenAI’s release · family 230Exact Hausdorff gauges for SLEOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: An exact Hausdorff gauge for SLE; An explicit exact Hausdorff gauge for SLE.

    • 2610.15066An Exact Hausdorff Measure as Natural Length for SLEDapeng Zhan · math.PR · 10 October 2026Main theorem takes its exact Hausdorff gauge theorem as input
  106. OpenAI’s release · family 241Rigidity of the Turing degreesOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: Rigidity of the Turing degrees.

    • 2610.16059Baire measurable functions and automorphisms of the truth-table degreesBjørn Kjos-Hanssen · math.LO · 12 October 2026Main theorem adapts its Turing-degree rigidity proof
  107. OpenAI’s release · family 252A torsion-free hyperbolic group that is neither residually finite nor linear over any fieldOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: A torsion-free hyperbolic group that is not residually finite.

    1 other paper cites it without following it up.

    • 2610.24836An example of a non-sofic hyperbolic groupHarjas Sandhu and Andreas Thom · math.GR · 17 October 2026Main construction builds on its non-residually-finite hyperbolic group, using its lemmas and sections
  108. OpenAI’s release · family 268The spin-one Haldane gapOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: The periodic spin-one Haldane gap; A boundary-field gap for the spin-one Heisenberg chain.

    • 2610.34330On the uniqueness of translation-invariant ground states of the S = 1 Heisenberg and J–D antiferromagnetic chainsHal Tasaki · math-ph · 25 October 2026Uniqueness of the infinite-chain ground state of the S = 1 chain, extending the machinery of its Haldane-gap manuscripts and using their numerical certificates
  109. OpenAI’s release · family 285Counterexamples to Baum–Connes and Kadison–KaplanskyOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: A torsion-free counterexample to reduced Baum–Connes injectivity; A torsion-free counterexample to the Kadison–Kaplansky projection conjecture; An irrational-trace counterexample to reduced Baum–Connes.

    • 2610.35674Projections in reduced group C*-algebras with prescribed trace values in (0,1)Guoliang Yu · math.OA · 26 October 2026Projections of every trace r ∈ (0,1) in reduced group C*-algebras, generalizing its construction of a projection of irrational trace
  110. OpenAI’s release · family 287 ★Isomorphism of the free group factorsOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: An isomorphism of the free group factors.

    2 other papers cite it without following it up.

    • 2610.19590On the Isomorphism of Semicircular AlgebrasD. Shlyakhtenko · math.OA · 14 October 2026Main theorem adapts its free group factor isomorphism proof to C*-algebras
  111. OpenAI’s release · family 291Cuntz comparison, nuclear dimension, and equivariant Jiang–Su stabilityOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Equivariant Jiang–Su Stability for Amenable Actions in the Unital Stably Finite Case; Cuntz comparison and Jiang–Su absorption; Nuclear dimension and Jiang–Su stability without elementary subquotients; Tracial projection methods and uniform property Gamma.

    1 other paper cites it without following it up.

    • 2611.00995Equivariant 𝒵-stability for amenable actionsEusebio Gardella · math.OA · 31 October 2026Extends its equivariant Jiang–Su stability theorem to amenable actions of arbitrary countable groups, by its Gaussian-variable technique
  112. OpenAI’s release · family 294Kaplansky's quasitrace conjecture and failure of tensor-product stable finitenessOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: A counterexample to Kaplansky's quasitrace conjecture and failure of tensor-product stable finiteness.

    1 other paper cites it without following it up.

    • 2610.27184Counterexamples to the Monotone Completeness ConjectureJananan Arulseelan and James E. Hanson · math.OA · 19 October 2026Main counterexample built by modifying its quasitrace counterexample, which the paper also reproves
  113. OpenAI’s release · family 298Two notions of free entropy differ even when both are finiteOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: A finite-entropy separation of microstates and nonmicrostates free entropy.

    • 2610.27243On counterexamples to the unification of free entropyDavid Jekel · math.OA · 19 October 2026Detailed presentation of a variant of its free-entropy example
  114. OpenAI’s release · family 302Radius of comparison equals half the mean dimensionOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Filtered products and boundary-preserving compression in complex cobordism; Radius of comparison equals half the mean dimension.

    • 2610.11143Mean dimension and radius of comparisonGeorge A. Elliott and Zhuang Niu · math.OA · 8 October 2026Main theorem extends its reverse bound; arguments based on the release preprint
  115. OpenAI’s release · family 308Finite Smith–Toda complexes at every heightOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Finite Smith–Toda Complexes at Varying Primes; A finite Smith–Toda complex V(4) at the prime 1009.

    • 2610.23215The Smith-Toda complex V(4) exists for p >= 11Vigleik Angeltveit · math.AT · 16 October 2026Main theorem improves its V(4) at p = 1009 to all p ≥ 11, following its strategy
  116. OpenAI’s release · family 324Lipschitz equivalent Banach spaces need not be linearly isomorphicOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Lipschitz Equivalent Separable Banach Spaces Need Not Be Linearly Isomorphic; Bi-Lipschitz Absorption of c0 Without a Linear Copy of c0.

    • 2611.06493Nonisomorphic separable Banach spaces at Lipschitz distance oneRyotaro Tanaka · math.FA · 4 November 2026Extends its construction to nonisomorphic separable Banach spaces at Lipschitz distance one
  117. OpenAI’s release · family 325The complete Crouzeix conjectureOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: A direct proof of the complete Crouzeix inequality; The complete Crouzeix theorem: optimal similarity and a common positive boundary representation.

    1 other paper cites it without following it up.

    • 2611.01503A general framework for universal bounds on the analytic functional calculusMihai Putinar and Georgios Tsikalas · math.FA · 1 November 2026Raises the two-step method of its direct proof to a general principle, recovering the constant 2 on convex domains
  118. OpenAI’s release · family 330A uniformly discrete counterexample to bounded approximation in Lipschitz-free spacesOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: A uniformly discrete counterexample to bounded approximation in Lipschitz-free spaces.

    • 2611.07930The Approximation Property without the Bounded Approximation Property in a Separable Banach Space and Its DualMa Yicen · math.FA · 5 November 2026Its renorming theorem is the input of the main theorem: X and X* with the approximation property but not the bounded one
  119. OpenAI’s release · family 336Spectral scalar curvature, Urysohn width, and macroscopic dimensionOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Spectral scalar curvature and uniform Urysohn width; Spectral scalar curvature and Urysohn width in dimension three; Positive scalar curvature and uniform codimension-two width.

    • 2610.28390A note on the cube inequalityYuchen Bi · math.DG · 20 October 2026Main theorem proved by adapting ideas of the manuscript
  120. OpenAI’s release · family 345Infinitely many closed geodesics on Riemannian spheres and closed three-manifoldsOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: Infinitely many closed geodesic images on every Riemannian sphere.

    • 2610.21054Infinitely many closed geodesic images on simply connected compact rank-one symmetric spacesZhiqi Chen, Shaoqiang Deng and Hui Zhang · math.DG · 15 October 2026Main theorem extends its sphere theorem to all CROSSes, taking its finite-cut and doubling strategy
  121. OpenAI’s release · family 347Counterexamples to stable-Morse and strong Arnold fixed-point boundsOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Hamiltonian Fixed Points Below the Stable Morse Number in Dimension Twenty-Two; Sharpness of the Cyclic Integral Floer Bound Below the Stable Morse Number; Hamiltonian Fixed Points Below the Stable Morse Number; A nondegenerate counterexample to the Morse-number Arnold bound; Three fixed points on the symplectic quadric threefold.

    • 2611.09694On the strong Arnold conjecture for GrassmanniansKenneth Blakey and Roman Krutowski · math.SG · 6 November 2026Generalizes the strategy of its counterexample to flag varieties: clean Hamiltonian diffeomorphisms with fewer fixed points than the strong Arnold conjecture predicts
  122. OpenAI’s release · family 349The Solomon–Yau least-volume conjectureOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: The Solomon–Yau least-volume theorem.

    1 other paper cites it without following it up.

    • 2610.20453Allen–Cahn and mean curvature flows from Clifford hypersurfaces to equatorsJingwen Chen · math.DG · 14 October 2026Main theorem extends the authors' flows to all dimensions using its Solomon–Yau resolution
  123. OpenAI’s release · family 365Joint metric and connection recovery from one boundary patchOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The family’s manuscripts: Determination of a metric and a unitary connection from one boundary patch; Smooth anisotropic uniqueness in the Calderón problem from one boundary patch; Nonuniqueness for bounded measurable scalar conductivities in three dimensions.

    3 other papers cite it without following it up.

    • 2611.10537Formal Proof of an Anisotropic Calderón Uniqueness TheoremCătălin I. Cârstea and Lauri Oksanen · math.AP · 7 November 2026A Lean 4 formalization of a special case of its Calderón uniqueness theorem
  124. OpenAI’s release · family 375De Giorgi's conjecture in dimension eightOpenAI, 6 October 2026community notes on oai-math-revised1follow-up paper

    The manuscript: A positive resolution of De Giorgi's conjecture in dimension eight.

    1 other paper cites it without following it up.

    • 2610.28080On the Allen–Cahn equationZhouyu Long and Wenming Zou · math.AP · 20 October 2026Main energy-growth theorem proved by following the approach of its De Giorgi manuscript
  125. OpenAI · Ten advances · result 8Ehrhart’s volume conjectureOpenAI, 1 August 2026the announcement · the Lean proofs1follow-up paper

    1 other paper cites it without following it up.

    • 2608.01040The equality case of Ehrhart's volume conjectureJihao Liu · math.CO · 2 August 2026Proves the equality case of Ehrhart's volume conjecture, with the frame and lemmas of its proof of the inequality

Every follow-up paper

The 236 papers, newest first, with the AI-generated work each one follows up. A paper following up two AI-generated works appears once for each; one following up several works of OpenAI’s October 6 release is listed under the first one it names.

Show the list
PaperFollows upWhat it does with it
2611.17543The Triangular Abrikosov Constant and the Thomas–Fermi Γ-Limit of Rotating Bose–Einstein CondensatesGiovanni Di Fratta and Valeriy V. Slastikov · math.AP · 12 November 2026OpenAI’s release, family 090: Triangular-lattice optimality, long-range Riesz and Coulomb energies, and spherical logarithmic energyIts Gaussian-minorant theorem gives the lower bound identifying the interaction coefficient of rapidly rotating Bose–Einstein condensates
2611.17124Why should ℤ be Diophantine in ℚ?Alexander Kozachinskiy, Hector Pasten and Xavier Vidaux · math.NT · 12 November 2026OpenAI’s release, family 004: Hilbert’s tenth problem over ℚA finitary version of its argument, conditional on finiteness of 2-primary Shafarevich–Tate groups, the abc conjecture and Dickson's conjecture
2611.16665Proof of the Quasi Riemann Hypothesis using ℚ[i]Xin Hang Ji · math.NT · 12 November 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisA shorter proof of a quasi-Riemann hypothesis for ζ (no zeros in Re s > 599/600), with quartic Gauss sums over ℤ[i] in place of its cubic ones
2611.16130A Diophantine definition of ℤ in ℚAshay Burungale · math.NT · 12 November 2026OpenAI’s release, family 004: Hilbert’s tenth problem over ℚProves ℤ Diophantine in ℚ by an approach inspired by its manuscript, taking its rank-one curve over ℚ(√2) and adapting its sieve argument
2611.14867Falconer's conjecture through descent and chainingWill Burstein, Alex Iosevich and Alexia Yavicoli · math.CA · 11 November 2026OpenAI’s release, family 073: The Falconer distance conjectureA complete proof of its theorem, following its architecture with modifications, and a stronger conclusion on the Fourier dimension of the distance set
2611.14837Two-Stage Mixed-Integer Quadratic Programming in Polynomial TimeAlberto Del Pia · math.OC · 10 November 2026Hidden Convexity via Symmetric Displacement Covers: Mixed-Integer Quadratic Programming and Integer Cubic Programming in Fixed Dimension (arXiv 2609.18266)Two-stage mixed-integer quadratic programs in polynomial time, generalizing its algorithm for mixed-integer quadratic programming in fixed dimension and using it for the subproblems
2611.14283Virtual Two-Step Nilpotency of the Fundamental Groups of Special Quasi-Projective VarietiesTianshu Jiang and Jiayu Li · math.AG · 10 November 2026OpenAI’s release, family 057: Fundamental groups of special complex varieties and root orbifoldsAdapts its proof of Campana's abelianity conjecture to special quasi-projective varieties: fundamental group virtually nilpotent of class at most two
2611.14131Euclidean Ramsey Theory for Lines and Finite ConfigurationsRyan Hoff, Gary Hu and Yumo Liu · math.CO · 10 November 2026OpenAI’s release, family 158: The Euclidean plane cannot be colored with five colorsIts theorem on rotation-invariant measures from the five-colour proof is an input to all its Euclidean Ramsey results
2611.13789On the proof of Falconer distance conjectureKotaro Inami, Jianbin Li, Yudong Liang, Bochen Liu and Francisco Romero Acosta · math.CA · 10 November 2026OpenAI’s release, family 073: The Falconer distance conjectureAn exposition, for graduate readers, of a proof following its main idea
2611.13697Gödel's Incompleteness Theorems in Grounded Arithmetican author of the workBryan Ford · math.LO · 10 November 2026Computable Quantification in Reflective Grounded Arithmetic (arXiv 2607.25533)A sequel by the same author: Gödel's incompleteness theorems for its system, resting on its soundness and consistency
2611.13697Gödel's Incompleteness Theorems in Grounded Arithmetican author of the workBryan Ford · math.LO · 10 November 2026Internalized Truth in Reflective Grounded Arithmetic (arXiv 2608.16140)A sequel by the same author: revises and extends its Gödel material, on its codes of terms and proofs and its proof checker
2611.13159Euclidean filling inequalities in CAT(0) spacesAlexander Lytchak · math.MG · 9 November 2026OpenAI’s release, family 337: Sharp Cartan–Hadamard isoperimetry and rigidityA streamlined version of its proof, extended to complete spaces, with rigidity in the equality case
2611.12756On the higher-order Titchmarsh divisor problemGuangwei Hu, Guangshi Lü, Zhiwei Wang and Jie Wu · math.NT · 9 November 2026OpenAI’s release, family 011: Prime-factor statistics of p-1Extends its Theorem 3.1 to every fixed nonzero shift, following its proof, for the asymptotic of Σ Λ(n)τₖ(n + h)
2611.12585A Large Gap between Galerkin Errors with and without Mesh-Uniform Regularity for the KdV Equation with Boundary Dissipationan author of the workKangsheng Liu · math.NA · 9 November 2026Linearize or Discretize First? Error Estimates under Mesh-Uniform Regularity and Exponential Stability for the KdV Equation with Boundary Dissipation (arXiv 2610.39269)A sequel by the same author: shows that its error bounds for the weighted Hermite–Galerkin scheme fail for the standard scheme
2611.12467An asymptotic formula for primes of the form u2+v2+1Yujiao Jiang, Yuk-Kam Lau and Guangshi Lü · math.NT · 9 November 2026OpenAI’s release, family 011: Prime-factor statistics of p-1Motohashi's conjecture on primes u² + v² + 1, with its harmonic-switching and high-moment method in the minor-arc bound
2611.12176From Frobenius seeds to self-orthogonal modules over symmetric algebrasHongxing Chen · math.RT · 9 November 2026OpenAI’s release, family 199: Counterexamples to Auslander–Reiten, Tachikawa and related homological conjecturesGeneralizes its counterexample to Tachikawa's second conjecture to a construction from any Frobenius algebra with a suitable module
2611.11566A rank-zero 2-converse to the Gross–Zagier and Kolyvagin theorem for elliptic curves with abelian two-division fieldAshay Burungale and Ye Tian · math.NT · 8 November 2026OpenAI’s release, family 006: Goldfeld’s conjecture: densities and mean analytic rankA self-contained proof, by its quadratic twist deformation method, that a finite 2^∞-Selmer group gives L(E,1) ≠ 0 when the two-division field is abelian
2611.11060The sharp Fourier extension estimate for the spherePang-hung Chung · math.AP · 8 November 2026OpenAI’s release, family 079: Local smoothing in three dimensionsThe full strong-type Fourier extension estimate for the sphere in every dimension n ≥ 3, from two of its estimates
2611.10537Formal Proof of an Anisotropic Calderón Uniqueness TheoremCătălin I. Cârstea and Lauri Oksanen · math.AP · 7 November 2026OpenAI’s release, family 365: Joint metric and connection recovery from one boundary patchA Lean 4 formalization of a special case of its Calderón uniqueness theorem
2611.10139The satisfiability conjecture for random k-SAT formulas. Human-friendly exposition of the recent OpenAI resolutionDavid Gamarnik · math.PR · 6 November 2026OpenAI’s release, family 235: Limiting random SAT thresholds, sharp variance and computabilityA reader-friendly exposition of its proof
2611.09697Geometric folding and maximal asymptotic varianceHaakan Hedenmalm · math.CV · 6 November 2026OpenAI’s release, family 072: Brennan's conjecture and the integral-means spectrumThe maximal asymptotic variance bound for the schlicht class, by a finite-scale form of the folding method that grew out of its manuscripts
2611.09694On the strong Arnold conjecture for GrassmanniansKenneth Blakey and Roman Krutowski · math.SG · 6 November 2026OpenAI’s release, family 347: Counterexamples to stable-Morse and strong Arnold fixed-point boundsGeneralizes the strategy of its counterexample to flag varieties: clean Hamiltonian diffeomorphisms with fewer fixed points than the strong Arnold conjecture predicts
2611.09386The exceptional set of Goldbach numbers among polynomial valuesI. Althöfer, D. Chen, A. Perelli and J. Pintz · math.NT · 6 November 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisMain theorem on 2φ(n) as a sum of two primes uses its 7/8 half-plane as the hypothesis of the fourth-moment estimates
2611.09361Positivity of metric entropy of the standard map at large parametersS. Slijepčević · math.DS · 6 November 2026OpenAI’s release, family 146: Positive metric entropy for the standard mapAn explicit entropy bound of order log k for the standard map, following its proof's architecture
2611.08819Reflexive polytopes can have doubly exponentially many verticesJohannes Hofscheier and Benjamin Nill · math.CO · 6 November 2026OpenAI’s release, family 025: Short Egyptian fractionsIts short Egyptian fractions, in Bloom and van Doorn's form, yield reflexive polytopes with doubly exponentially many facets
2611.08268A Note on the OpenAI Construction of Counterexamples to the Kaplansky Direct-Finiteness ConjectureAndrei Jaikin-Zapirain · math.GR · 5 November 2026OpenAI’s release, family 197: A torsion-free group algebra that is not directly finiteAn exposition of its construction, adapted to every prime p: hyperbolic groups whose group algebra over 𝔽_p is not directly finite
2611.08224Fourier inequalities and sign uncertainty IIRoni Edwin · math.CA · 5 November 2026OpenAI, Ten advances, result 1: High-dimensional sphere packingAdapts the argument of its Chapter 1 (the sign-uncertainty asymptotics) to an upper bound for the best constant in ‖f‖₂² ≤ w_d ‖f‖₁ ‖f̂‖₁ for radial functions
2611.08198A diameter-one set in ℝ5 that cannot be covered by six diameter-one ballsA. Arman, A. Bondarenko, A. Prymak and D. Radchenko · math.MG · 5 November 2026OpenAI’s release, family 156: Borsuk's conjecture fails in dimension nineIts nine-dimensional Borsuk construction carried to ℝ⁵: a set of diameter 1 not covered by six balls of diameter 1
2611.07930The Approximation Property without the Bounded Approximation Property in a Separable Banach Space and Its DualMa Yicen · math.FA · 5 November 2026OpenAI’s release, family 330: A uniformly discrete counterexample to bounded approximation in Lipschitz-free spacesIts renorming theorem is the input of the main theorem: X and X* with the approximation property but not the bounded one
2611.07085The square function estimates for cones in arbitrary dimensions and elliptic quadricsFaruk Temur · math.CA · 4 November 2026OpenAI’s release, family 079: Local smoothing in three dimensionsReverse square function estimates for the cone and for spheres and other surfaces, from its local smoothing statements taken as hypotheses
2611.06869Finite-Time Blowup for Navier-Stokes with Smooth Forcing, Part II: Residual Correction via Oscillatory PulsesZhen Lei and Xiao Ren · math.AP · 4 November 2026OpenAI: Finite time blowup for Navier–StokesThe second of two papers expounding its pulse corrections, claiming no original contribution
2611.06493Nonisomorphic separable Banach spaces at Lipschitz distance oneRyotaro Tanaka · math.FA · 4 November 2026OpenAI’s release, family 324: Lipschitz equivalent Banach spaces need not be linearly isomorphicExtends its construction to nonisomorphic separable Banach spaces at Lipschitz distance one
2611.06278Uniqueness of the hexagonal lattice's universal optimalityMarkus Faulhuber and Philipp Petersen · math.MG · 4 November 2026OpenAI’s release, family 090: Triangular-lattice optimality, long-range Riesz and Coulomb energies, and spherical logarithmic energyUniqueness: by its auxiliary functions, the triangular lattice is the only periodic configuration minimizing every Gaussian energy
2611.05807Near-Optimal Complexity of Finding Approximate Local Minima of Nonconvex Finite SumsQihao Zhou · math.OC · 4 November 2026Matching Lower Bounds for Randomized First-Order Methods in Hessian-Lipschitz Nonconvex Optimization (arXiv 2610.00245)Replicates its spectral hard instance across components for the main theorem's lower bound against unrestricted randomized algorithms
2611.05534One-third Hölder stability of optimal transport maps from log-concave sourcesSongbo Wang and Xiaozhen Wang · math.AP · 3 November 2026OpenAI’s release, family 374: Sharp one-third stability of Brenier mapsExtends its one-third Hölder stability of Brenier maps to log-concave sources
2611.05452On Fujita's Advanced Generalized Mukai ConjectureDuo Li · math.AG · 3 November 2026OpenAI’s release, family 063: The generalized Mukai conjectureFujita's advanced generalized Mukai conjecture, from its generalized Mukai inequality and its Gromov–Witten correspondences
2611.04599Kaplansky's conjectures fail for torsion-free hyperbolic groups in positive characteristicNidhal Mghirbi and Seth Douglas · math.GR · 3 November 2026OpenAI’s release, family 197: A torsion-free group algebra that is not directly finiteProves its torsion-free groups hyperbolic and adapts its construction to every odd characteristic, so that all four Kaplansky conjectures fail for hyperbolic groups
2611.04297Fractional colorings of Kr-free graphs with local demandsEwan Davies, Abhishek Dhawan, Abhishek Methuku, Huy Nguyen and Jaehyeon Seo · math.CO · 3 November 2026OpenAI’s release, family 184: Correspondence coloring with a fixed forbidden subgraphThe Kelly–Postle conjecture on fractional colouring with local demands, by adapting its triangle-weight argument
2611.03572Counterexamples to the Jacobian Conjecture from the viewpoint of lineabilityL. Bernal-González, M. C. Calderón-Moreno and J. A. Prado-Bassas · math.AG · 2 November 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')Studies the size of the family of counterexamples in dimension N ≥ 3, nonempty through it: lineability exactly 1, affine coneability, semigroups under composition
2611.03515The proof of the Harary–Hill conjecture: a detailed exposition and an extension to rotation systemsGelasio Salazar · math.CO · 2 November 2026OpenAI’s release, family 165: The Harary–Hill and Zarankiewicz crossing-number formulasA detailed exposition of its proof of the Harary–Hill conjecture for odd n, extending its polynomial method to rotation systems
2611.03225Human Variations on AI MathematicsJunyan Cao and Mihai Paun · math.AG · 2 November 2026OpenAI’s release, family 034: Log abundance for compact Kähler spaces under logarithmic Iitaka subadditivityAn extension theorem for pluricanonical sections by the method of its minimal-metrics manuscript, giving another proof of its log abundance theorem
2611.03214The Dirichlet problem for quaternionic Monge–Ampère equations on flat quaternionic quotientsXi-Nan Ma and Jingyong Zhu · math.AP · 2 November 2026OpenAI’s release, family 079: Local smoothing in three dimensionsQuaternionic Monge–Ampère Dirichlet problems on quotients of ℍ^m, of which the periodic theorem it uses for its local smoothing estimates is a special case
2611.02531No Borel upper bound for the Fσ equivalence relationsan author of the workRuijun Wang · math.LO · 2 November 2026There is no universal Fσ equivalence relation (arXiv 2610.31052)A sequel by the same author: no Borel equivalence relation of any complexity bounds all F_σ equivalence relations, strengthening its theorem by another proof
2611.02454The geometric case of the Erdős Similarity Conjecture. A human exposition of the OpenAI proofAlex Iosevich and Alexia Yavicoli · math.PR · 2 November 2026OpenAI’s release, family 084: The geometric case of the Erdős similarity conjectureA detailed exposition of its proof, related to earlier random-cell arguments and derandomized
2611.02448The polyharmonic Hénon–Lane–Emden conjecturePhuong Le · math.AP · 2 November 2026OpenAI’s release, family 370: The Lane–Emden and Hénon–Lane–Emden conjecturesExtends its argument to integral systems with Riesz kernels, proving the polyharmonic Hénon–Lane–Emden conjecture
2611.01781Inhomogeneous Fiber Refinement of Gallagher's TheoremQian Li, Bo Tan and Qing-Long Zhou · math.NT · 1 November 2026OpenAI’s release, family 022: The weak inhomogeneous Duffin–Schaeffer conjectureIts theorem gives the divergence half of the main theorem on weighted inhomogeneous sets
2611.01745A remark on the AI proof of the Deligne–Drinfeld conjectureNikita Markarian · math.AG · 1 November 2026OpenAI’s release, family 008: The Deligne–Drinfeld conjectureA geometric interpretation of the injective projection in its characteristic-two proof, making its reconstruction explicit
2611.01610Logarithmic Sobolev and exponential inequalities for submanifolds in Cartan–Hadamard manifoldsHongyu Chen and Li Ma · math.AP · 1 November 2026OpenAI’s release, family 337: Sharp Cartan–Hadamard isoperimetry and rigidityExtends its Michael–Simon type inequality to every codimension, for a logarithmic Sobolev inequality on submanifolds
2611.01532The massive phase of the two-dimensional Heisenberg model is stable under small anisotropyYan Ru Pei · math-ph · 1 November 2026OpenAI’s release, family 215: Canonical O(3) continuum limit and exact O(4) mass asymptoticsAdapts its proof, with its annulus estimate as input, to show that exponential decay survives small anisotropy of the couplings
2611.01503A general framework for universal bounds on the analytic functional calculusMihai Putinar and Georgios Tsikalas · math.FA · 1 November 2026OpenAI’s release, family 325: The complete Crouzeix conjectureRaises the two-step method of its direct proof to a general principle, recovering the constant 2 on convex domains
2611.01490No infinite cluster at criticality for long-range percolation with fast-decaying interactionsYan Ru Pei · math.PR · 1 November 2026OpenAI’s release, family 213: Critical percolation on every quasi-transitive graphAdapts its proof of no percolation at criticality on ℤ³ to long-range percolation on ℤᵈ
2611.01423The classification of Fano contact manifolds revisitedBaohua Fu · math.AG · 1 November 2026OpenAI’s release, family 062: Projective contact classification and the LeBrun–Salamon conjectureAn account, with simplifications, of its proof that every Fano contact manifold is an adjoint variety
2611.01338Threshold existence for random NAE-k-SAT and random hypergraph 2-colourabilityPavel Zakharov · math.CO · 31 October 2026OpenAI’s release, family 235: Limiting random SAT thresholds, sharp variance and computabilityAdapts its approach from random k-SAT to limiting thresholds for random NAE-k-SAT and hypergraph 2-colourability
2611.00995Equivariant 𝒵-stability for amenable actionsEusebio Gardella · math.OA · 31 October 2026OpenAI’s release, family 291: Cuntz comparison, nuclear dimension, and equivariant Jiang–Su stabilityExtends its equivariant Jiang–Su stability theorem to amenable actions of arbitrary countable groups, by its Gaussian-variable technique
2611.00176Coinduction in Grounded Arithmetican author of the workBryan Ford · math.LO · 30 October 2026Computable Quantification in Reflective Grounded Arithmetic (arXiv 2607.25533)A sequel by the same author: its system, through the inductive extension, with a coinduction rule and a dichotomy rule, with soundness and consistency
2611.00176Coinduction in Grounded Arithmetican author of the workBryan Ford · math.LO · 30 October 2026Idealizing Useful Fictions in Omega Grounded Arithmetic (arXiv 2608.17862)A sequel by the same author: one of the new systems contains it, and a main theorem derives its omniscience rule from the dichotomy rule
2610.41931A Tropical Viewpoint on the Nagata–Biran–Szemberg ConjectureJuliette Bruce · math.AG · 30 October 2026OpenAI’s release, family 039: Nagata’s conjecture and maximal Seshadri constantsIts Nagata–Biran–Szemberg theorem with an explicit threshold
2610.41766An odd version of the Grothendieck–Teichmüller Lie algebra and the odd Deligne–Drinfeld conjectureFlorian Naef and Thomas Willwacher · math.QA · 30 October 2026OpenAI’s release, family 008: The Deligne–Drinfeld conjectureAn odd analogue of its Deligne–Drinfeld theorem
2610.41286On the Turán number for the Cartesian product of an odd cycle and a treeYa-Hong Chen, Yong Lu, Quan Sun, Qi Wu and Long-Tu Yuan · math.CO · 30 October 2026A Counting Proof for Erdős Problem 548The Turán number of the product of an odd cycle and a tree, with the Erdős–Sós theorem it proved as an input
2610.40811Integral Hodge structures and atoms of fourfoldsJérémy Guéré · math.AG · 29 October 2026OpenAI’s release, family 054: Irrational cubic fourfolds with Hodge-theoretic and categorical K3 associationsA new proof of a theorem of its manuscript on irrational cubic fourfolds
2610.40439Shifted convolutions of divisor functionsMiguel N. Walsh · math.NT · 29 October 2026OpenAI’s release, family 011: Prime-factor statistics of p-1The asymptotic for shifted convolutions of divisor functions
2610.39272Sharp potential and Brenier map W2-stability for log-concave sourcesYuanlong Ruan · math.AP · 29 October 2026OpenAI’s release, family 374: Sharp one-third stability of Brenier mapsExtends its one-third stability of Brenier maps to strongly log-concave sources
2610.39102Logarithmic transformations on the six-sphereJeff Viaclovsky · math.AG · 29 October 2026The (3,4,∞) modular family of 2-tori, completed at its three special points, is a complex structure on S⁶Infinitely many choices of the multiplicities in its construction give complex structures on S⁶, pairwise biholomorphically inequivalent
2610.39096Sparse domination and an extended Lp range for the trilinear Hilbert transformBingyang Hu · math.CA · 29 October 2026OpenAI’s release, family 086: An L3 bound for the trilinear Hilbert transformSparse bounds for the trilinear Hilbert transform from its local estimate and stopping-time argument, extending its L³ × L³ × L³ → L¹ bound
2610.38736Transverse rational curves and the homogeneity of contact Fano manifoldsZhengyu Hu, Kefeng Liu and Xueyuan Wan · math.AG · 28 October 2026OpenAI’s release, family 062: Projective contact classification and the LeBrun–Salamon conjectureHomogeneity of every contact Fano manifold, which contains its classification, with a construction adapted from its manuscript
2610.38658Larger Crossing FamiliesJoachim Orthaber and Patrick Schnider · math.CO · 28 October 2026Linear crossing families via dual levels (arXiv 2609.29449)Improves its bound for crossing families in line-separated sets, giving crossing families of size about n/79
2610.37817The generalized Cartan-Hadamard conjectureMarcos Agnoletto, Márcio Fabiano Da Silva, Stefano Nardulli and Reinaldo Resende · math.DG · 27 October 2026OpenAI’s release, family 337: Sharp Cartan–Hadamard isoperimetry and rigidityThe generalized Cartan–Hadamard conjecture with rigidity, adapting its Michael–Simon inequality and balancing maps
2610.36588Square-root logarithmic bounds for the arithmetic complexity of factorial multiplesLiang Zhang · math.NT · 27 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisMain theorem on the arithmetic complexity of multiples of k! conditional on a fixed zero-free half-plane
2610.35760Endpoint gradient bounds for centered convex maximal operators and probability kernelsZhouyu Long · math.CA · 26 October 2026OpenAI’s release, family 085: Endpoint Sobolev regularity of centered disk averagesExtends its planar endpoint gradient bound for the centered disk maximal operator to convex bodies in every dimension, adapting its method
2610.35674Projections in reduced group C*-algebras with prescribed trace values in (0,1)Guoliang Yu · math.OA · 26 October 2026OpenAI’s release, family 285: Counterexamples to Baum–Connes and Kadison–KaplanskyProjections of every trace r ∈ (0,1) in reduced group C*-algebras, generalizing its construction of a projection of irrational trace
2610.34548Inductive Grounded Arithmetic: Goodstein, the Hydra, and Γ0an author of the workBryan Ford · math.LO · 25 October 2026Computable Quantification in Reflective Grounded Arithmetic (arXiv 2607.25533)A sequel by the same author: the system with an added induction rule, and its soundness, consistency and ordinal results
2610.34444Proof of the Harary–Hill conjecture: an expositionMartin Balko · math.CO · 25 October 2026OpenAI’s release, family 165: The Harary–Hill and Zarankiewicz crossing-number formulasA simplified exposition of its proof of the Harary–Hill conjecture
2610.34396A note on consequences of Mahler's inequalitiesYanlu Lian and Fei Xue · math.MG · 25 October 2026OpenAI’s release, family 087: The Mahler conjectures, functional inequalities and polar-product symplectic widthMakai's inequality in every dimension, derived from its symmetric Mahler theorem
2610.34330On the uniqueness of translation-invariant ground states of the S = 1 Heisenberg and J–D antiferromagnetic chainsHal Tasaki · math-ph · 25 October 2026OpenAI’s release, family 268: The spin-one Haldane gapUniqueness of the infinite-chain ground state of the S = 1 chain, extending the machinery of its Haldane-gap manuscripts and using their numerical certificates
2610.34125The effective mass of the Pauli-Fierz electron: a cutoff-independent lower bound and the breakdown of mass renormalizationan author of the workShinichiro Futakuchi · math-ph · 25 October 2026The sharp leading asymptotics of the self-energy of the spinless Pauli-Fierz model (arXiv 2610.34122)Takes its sharp self-energy asymptotics as the input of one of its two main theorems, a lower bound of order λ^(8/7−ε) on the growth of the effective-mass ratio (the same author)
2610.34036A simple construction of smooth counterexamples to the Loewner and Carathéodory conjecturesNaoya Ando and Masaaki Umehara · math.DG · 25 October 2026Smooth counterexamples to the Loewner and Carathéodory conjectures (X post)A simpler construction of its smooth counterexamples, by a variant of it with a trigonometric seed
2610.33900A remark on the global minimality of the cracktipan author of the workFrancesco Deangelis, Camillo De Lellis and Matteo Focardi · math.AP · 24 October 2026Solution of the Mumford-Shah conjecture (arXiv 2609.26732)A short proof that the cracktip is a global minimizer, with its resolution of the Mumford–Shah conjecture as the interior regularity input (shares an author with it)
2610.32920Hot Spots: From Smooth to Nonsmooth Simply Connected Planar DomainsRodrigo Banñuelos · math.AP · 23 October 2026OpenAI’s release, family 369: The hot spots conjecture for simply connected planar domainsThe hot-spots conclusion on simply connected domains of finite area, with its theorem for smooth domains as hypothesis
2610.32789Triangle-Free Sets in Exact-Distance Hamming GraphsBence Csajbók, Padmini Mukkamala and Zoltán Lóránt Nagy · math.CO · 23 October 2026OpenAI, Ten advances, result 9: Multicolor Ramsey numbersIts superexponential multicolour Ramsey lower bound as the input of the main ternary construction of triangle-free sets in Hamming graphs
2610.31631Nonvanishing of 2-primary Kolyvagin classes via simultaneous quadratic twistsZhiqi Sun · math.NT · 22 October 2026On The Manin Constant (arXiv 2610.25672)Uses its theorem that the Manin constant of an optimal elliptic curve over ℚ is ±1 to normalize the modular parametrization in its main nonvanishing criterion for 2-primary Kolyvagin classes
2610.31631Nonvanishing of 2-primary Kolyvagin classes via simultaneous quadratic twistsZhiqi Sun · math.NT · 22 October 2026OpenAI’s release, family 002: The full BSD formula from low Selmer corankMain index formula and nonvanishing theorem use its 2-converse and 2-primary BSD formula as inputs throughout
2610.31402Landau-Siegel zeros over a base field and effective Brauer-SiegelWilliam D. Banks · math.NT · 22 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisMain theorems extend (Kerr's form of) its argument excluding Landau–Siegel zeros to a base field
2610.31167Liouville Theorems and Sharp Homogeneity Defects for Weighted Potential SystemsShengbing Deng and Bo Zhang · math.AP · 22 October 2026OpenAI’s release, family 370: The Lane–Emden and Hénon–Lane–Emden conjecturesMain Liouville theorem proved within its source-layer interval-pressure framework
2610.31112Irrationality Exponents of Real CM Elliptic Periods and Powers of the Lemniscate PeriodGregory V. Chudnovsky and David V. Chudnovsky · math.NT · 22 October 2026OpenAI’s release, family 017: The irrationality exponent of π is 2Main theorem proved by its separated-weight method for μ(π) = 2, partly formalized with its Lean library
2610.30976Rigidity, flexibility, and a sign errorYanki Lekili and Nick Sheridan · math.SG · 22 October 2026OpenAI’s release, family 032: Hodge and Kuga–Satake results for all projective K3 surfacesThe whole paper examines its withdrawn Hodge-conjecture manuscripts and locates their sign error
2610.30797OpenAI partially disproves Littlewood's 1961 Lα-flat theoremel Houcein el Abdalaoui · math.NT · 22 October 2026OpenAI’s release, family 076: Real ultraflat Littlewood polynomials and unbounded binary merit factorsPresents and analyses its counterexample to Littlewood's theorem
2610.30490A note on the proof of Barnette's conjectureZhentao Li · math.CO · 21 October 2026OpenAI’s release, family 180: Barnette’s Hamiltonian-cycle conjectureExposition of its proof of Barnette's conjecture
2610.30476An explicit elementary proof of no Landau–Siegel zerosVictor Zhenyu Guo · math.NT · 21 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisMain theorem is an explicit elementary version following its Frobenius method for Landau–Siegel zeros
2610.30063On the proof of the Kozma-Nitzan conjectureJean-Baptiste Gouéré · math.PR · 21 October 2026OpenAI’s release, family 213: Critical percolation on every quasi-transitive graphExposition of its proof of the Kozma–Nitzan conjecture
2610.30050On OpenAI's proof of the Nagata conjectureJenia Tevelev · math.AG · 21 October 2026OpenAI’s release, family 039: Nagata’s conjecture and maximal Seshadri constantsExpository notes on its proof of Nagata's conjecture
2610.30027The recent counterexamples to the Eisenbud-Schreyer-Weyman Ulrich problem (2003) and to Hochster's small Cohen-Macaulay conjecture (1973): a survey, with a generalization to arbitrary dimensionCristian Anghel · math.AG · 21 October 2026OpenAI’s release, family 195: A counterexample to the small Cohen–Macaulay module conjectureSurvey expounding its small Cohen–Macaulay counterexample, whose new results generalize its argument to every dimension
2610.29470The exceptional set of Goldbach numbers within the set of squaresA. Perelli and J. Pintz · math.NT · 21 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisMain theorem's major-arc evaluation uses its 7/8 zero-free half-plane as an input
2610.28478A single generation theorem for simple unital C*-algebras with CPoU and comparisonXin Ma · math.OA · 20 October 2026OpenAI’s release, family 296: The generator problem for finite factorsMain theorem assumes its two-generation theorem for II₁ factors
2610.28473C*-selflessness and 𝒵-stability for Thompson's group FChris Bruce and Shirly Geffen · math.OA · 20 October 2026OpenAI’s release, family 248: Thompson's group F is nonamenableBoth main theorems presented as applications of its nonamenability of F and its Corollary 1.3 on infinite tensor powers
2610.28390A note on the cube inequalityYuchen Bi · math.DG · 20 October 2026OpenAI’s release, family 336: Spectral scalar curvature, Urysohn width, and macroscopic dimensionMain theorem proved by adapting ideas of the manuscript
2610.28080On the Allen–Cahn equationZhouyu Long and Wenming Zou · math.AP · 20 October 2026OpenAI’s release, family 375: De Giorgi's conjecture in dimension eightMain energy-growth theorem proved by following the approach of its De Giorgi manuscript
2610.28079Irrationality exponents of logarithmic numbers: Values, quotients, and graded approximationGregory V. Chudnovsky · math.NT · 20 October 2026OpenAI’s release, family 017: The irrationality exponent of π is 2Main bounds obtained by adapting its interpolation and determinant method for the irrationality exponent of π
2610.27932Thompson's group F is not weakly amenableKoji Fujiwara · math.GR · 20 October 2026OpenAI’s release, family 248: Thompson's group F is nonamenableMain theorem deduced from its theorem that F is nonamenable
2610.27875Classification of positive solutions to the critical Lane-Emden systemYuxia Guo, Yichen Hu, Shaolong Peng and Tingfeng Yuan · math.AP · 20 October 2026OpenAI’s release, family 370: The Lane–Emden and Hénon–Lane–Emden conjecturesMain classification proved by its source-layer interval comparison method
2610.27243On counterexamples to the unification of free entropyDavid Jekel · math.OA · 19 October 2026OpenAI’s release, family 298: Two notions of free entropy differ even when both are finiteDetailed presentation of a variant of its free-entropy example
2610.27184Counterexamples to the Monotone Completeness ConjectureJananan Arulseelan and James E. Hanson · math.OA · 19 October 2026OpenAI’s release, family 294: Kaplansky's quasitrace conjecture and failure of tensor-product stable finitenessMain counterexample built by modifying its quasitrace counterexample, which the paper also reproves
2610.26835Zero-free half-planes for Hecke L-functions over number fieldsBaiying Liu · math.NT · 19 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisExtends its cubic-theta argument for the 7/8 half-plane to Hecke L-functions, using the primitive-roots manuscript's constructions
2610.26832Controllability of polynomial systems is undecidable and not recursively enumerablean author of the workJun Liu and Ivan O. Shevchenko · math.OC · 19 October 2026Global asymptotic stability of homogeneous polynomial vector fields is undecidable (arXiv 2610.12434)Takes its Σ⁰₁-completeness of global asymptotic stability as the input of its main theorem, that global controllability of polynomial control systems is undecidable and not recursively enumerable (two of the same authors)
2610.26756Fourth moments of prime exponential sums and exceptional sets for sums and differences of primesHuixi Li and Zhixin Liu · math.NT · 19 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisMain theorem conditional on a quasi-Riemann hypothesis with Θ < 11/12, supplied by its 7/8 half-plane
2610.26656Weighted potential and Green systems with Dini coefficients and primitive growthHongtao Hu, Zhouyu Long and Wenming Zou · math.AP · 19 October 2026OpenAI’s release, family 370: The Lane–Emden and Hénon–Lane–Emden conjecturesProof of the main nonexistence theorem inherits and adapts its interval-pressure framework
2610.26539The irrationality exponent of the lemniscate period: A determinant proof with a Lean 4 formalizationGregory V. Chudnovsky · math.NT · 19 October 2026OpenAI’s release, family 017: The irrationality exponent of π is 2Main theorem proved by its π-exponent method transposed to the lemniscate, formalized on 828 modules of its Lean library
2610.26419A connection inequality for the random-cluster modelJianping Jiang, Sike Lang and Yuxuan Sun · math.PR · 19 October 2026θ(p_c) = 0 for Bernoulli bond percolation on ℤ^d in all dimensions d ≥ 2, in Lean 4 / MathlibExtends the Kozma–Nitzan inequality, which it says AI proved first, to random-cluster measures with q ≥ 1, through Blanc-Renaudie's human version of the proof
2610.26419A connection inequality for the random-cluster modelJianping Jiang, Sike Lang and Yuxuan Sun · math.PR · 19 October 2026Kozma–Nitzan Connection Inequalities for Bernoulli PercolationExtends the Kozma–Nitzan inequality, which it says AI proved first, to random-cluster measures with q ≥ 1, through Blanc-Renaudie's human version of the proof
2610.26388Pointwise convergence of bilinear ergodic averages for commuting transformationsZhuowen Guo, Rongzhong Xiao and Shuhao Zhang · math.DS · 19 October 2026OpenAI’s release, family 082: Annular variation and dyadic absolute bounds for the triangular Hilbert transformProof of the main theorem adapts its matrix-energy and Gaussian heat-flow method
2610.25531Regularity of asymptotically axisymmetric Navier-Stokes flows with rotationally invariant curl of the forceBenjamin Ingimarson, Igor Kukavica and Qi Xu · math.AP · 18 October 2026Regularity of asymptotically axisymmetric solutions to the 3D Navier-Stokes equations with analytic forcing (arXiv 2609.20803)Answers the question of its Remark 2.5
2610.25357Studying proof data with proof theory: the case of AI-generated mathematicsSafal Raman Aryal · math.LO · 18 October 2026OpenAI’s release as a wholeMain measurements are made on 96 of its Lean projects as the principal corpus of AI-written proofs (the release studied as data, not used mathematically)
2610.24900All totally real abelian number fields of class number one are EuclideanHarald Grobner · math.NT · 17 October 2026OpenAI’s release, family 029: Primitive roots for every admissible integer baseMain theorem takes its zero-free region and its Proposition 2.1 as black-box inputs
2610.24889Alpöge's counterexample to the JC, Economic Stabilization, and Zhao's Vanishing ConjectureMichiel de Bondt · math.AG · 17 October 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')Counterexamples to the homogeneous Jacobian conjecture in dimension 15 and to Zhao's vanishing conjecture, by economic stabilization of a tweaked version of it
2610.24836An example of a non-sofic hyperbolic groupHarjas Sandhu and Andreas Thom · math.GR · 17 October 2026OpenAI’s release, family 252: A torsion-free hyperbolic group that is neither residually finite nor linear over any fieldMain construction builds on its non-residually-finite hyperbolic group, using its lemmas and sections
2610.24612Human Exposition of Claude's 2/3 Theorem on the zeros of zetaJ. Arias de Reyna · math.NT · 17 October 2026More than two thirds of the zeta zeros are simple and on the critical line (arXiv 2608.13637)A self-contained exposition of its proof, with a modification the author says the argument needs
2610.24080An equiregular counterexample to the Sard conjecture for sub-Riemannian endpoint mapsBangjie Wang and Igor Zelenko · math.DG · 16 October 2026A sub-Riemannian endpoint map with no regular values (arXiv 2609.05991)Answers the question its counterexample to the sub-Riemannian Sard conjecture left open
2610.23853Brennan's Conjecture and the Universal Integral Means SpectrumIlia Binder, Oleg Ivrii, Istvan Prause and Stanislav Smirnov · math.CV · 16 October 2026OpenAI’s release, family 072: Brennan's conjecture and the integral-means spectrumA shorter proof of a stronger form of its Brennan theorem, found while checking it
2610.23719Confluence relations, Hoffman basis over ℤ(2), and the Deligne–Drinfeld conjectureMinoru Hirose · math.NT · 16 October 2026OpenAI’s release, family 008: The Deligne–Drinfeld conjectureAbstract and title present the main theorem as yielding an alternative proof of its Deligne–Drinfeld theorem
2610.23215The Smith-Toda complex V(4) exists for p >= 11Vigleik Angeltveit · math.AT · 16 October 2026OpenAI’s release, family 308: Finite Smith–Toda complexes at every heightMain theorem improves its V(4) at p = 1009 to all p ≥ 11, following its strategy
2610.22927Unconditional Bounds for Two, Three and Four Consecutive Moderate Gaps between Zeros of the Riemann Zeta FunctionCésar Angón, Erick Ross, Ze Sen Tang and Boyang Zhao · math.NT · 16 October 2026More than two thirds of the zeta zeros are simple and on the critical line (arXiv 2608.13637)Adapts its method to consecutive moderate gaps between zeros
2610.22802The fractional Hénon-Lane-Emden conjecture and some critical nonexistence phenomenaKui Li, Haoyang Lu and Zhitao Zhang · math.AP · 16 October 2026OpenAI’s release, family 370: The Lane–Emden and Hénon–Lane–Emden conjecturesMain theorem refines and extends its Hénon–Lane–Emden method to fractional orders
2610.22612Proof Perplexity: A Probabilistic Approach to Measuring non-triviality in Machine Generated ProofsMadhuparna Das and Soumya Banerjee · math.LO · 15 October 2026More than two thirds of the zeta zeros are simple and on the critical line (arXiv 2608.13637)A measure of the non-triviality of formal proofs, applied only to its Lean formalization
2610.22536Failure of direct finiteness for group algebras in every odd characteristicElena I. Bunina, Winston Lee, Alexei Miasnikov and Kyrylo Muliarchyk · math.GR · 15 October 2026OpenAI’s release, family 196: A counterexample to Kaplansky’s zero-divisor conjectureAdapts the bounded-path and planar-localization arguments of its counterexample over 𝔽₂ to every odd prime p
2610.22222A quantitative form of the planar Falconer distance theoremGuozheng Cheng, Xiang Fang, Menghan Li, Hongdou Qu and Chengbo Xiao · math.CA · 15 October 2026OpenAI’s release, family 073: The Falconer distance conjectureMain theorem develops its multiscale Falconer method into a quantitative form
2610.21511Imaginary quadratic fields of class number at most 2000Iwao Kimura · math.NT · 15 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisMain result rests on making its Landau–Siegel argument explicit, with a Lean formalization
2610.21263The Rate of the Delsarte Bound for Binary Codes and Spheresan author of the workAndrew Salmon · math.CO · 15 October 2026The half-rate linear programming bound for binary codes is 1/2−1/π (arXiv 2609.03736)Extends its rate of Delsarte's bound to every distance and to spherical codes (the same author)
2610.21263The Rate of the Delsarte Bound for Binary Codes and SpheresAndrew Salmon · math.CO · 15 October 2026OpenAI, Ten advances, result 2: Binary and spherical codesExtends its constructions to find the exact rate of Delsarte's bound for binary and spherical codes
2610.21099A simpler proof that there are no Landau–Siegel zerosBryce Kerr · math.NT · 15 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisSimplified proof of its Landau–Siegel theorem
2610.21054Infinitely many closed geodesic images on simply connected compact rank-one symmetric spacesZhiqi Chen, Shaoqiang Deng and Hui Zhang · math.DG · 15 October 2026OpenAI’s release, family 345: Infinitely many closed geodesics on Riemannian spheres and closed three-manifoldsMain theorem extends its sphere theorem to all CROSSes, taking its finite-cut and doubling strategy
2610.2069186 unknowns for Hilbert's tenth problem over the rationalsZhi-Wei Sun · math.NT · 14 October 2026OpenAI’s release, family 004: Hilbert’s tenth problem over ℚMain theorem refines its negative solution of Hilbert's tenth problem over Q to 86 unknowns
2610.20496Smoothness criterion for Fano orbifolds and bases of Lagrangian fibrationsZhiyu Liu · math.AG · 14 October 2026OpenAI’s release, family 041: Hyperkähler SYZ and projective-space basesMain theorem is a simpler proof of the second part of its result, combining its ideas with the author's
2610.20453Allen–Cahn and mean curvature flows from Clifford hypersurfaces to equatorsJingwen Chen · math.DG · 14 October 2026OpenAI’s release, family 349: The Solomon–Yau least-volume conjectureMain theorem extends the authors' flows to all dimensions using its Solomon–Yau resolution
2610.20242On the Hardy–Littlewood–Chowla conjectureGuangshi Lü, Zhiwei Wang and Jie Wu · math.NT · 14 October 2026OpenAI’s release, family 011: Prime-factor statistics of p-1Type II estimate follows its determinant and memory method; root-counting lemma reproved from it
2610.20058Primes between 8th powers?Matt Visser · math.NT · 14 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisAll main results conditional on its 7/8 half-plane
2610.19590On the Isomorphism of Semicircular AlgebrasD. Shlyakhtenko · math.OA · 14 October 2026OpenAI’s release, family 287: Isomorphism of the free group factorsMain theorem adapts its free group factor isomorphism proof to C*-algebras
2610.18491Some applications of the quasi-Riemann hypothesis to class groups and small generators of number fieldsChristopher Frei, Efthymios Sofos and Martin Widmer · math.NT · 13 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisMain results conditional on the quasi-Riemann hypothesis, made unconditional by the release
2610.18360A note on the proof of nonamenability of FSang-hyun Kim and Yash Lodha · math.GR · 13 October 2026OpenAI’s release, family 248: Thompson's group F is nonamenableExposition of its proof
2610.17891The logarithmic exponent of the Ramsey number r(4,t)Qizhong Lin and Qiuyang Zhang · math.CO · 13 October 2026OpenAI’s release, family 170: Sharp logarithmic exponents for off-diagonal Ramsey numbersMain lower bound takes its r(5,t) bound as external input
2610.17662Zero divisors in group algebras of torsion-free groups in odd characteristicS. V. Ivanov · math.GR · 13 October 2026OpenAI’s release, family 196: A counterexample to Kaplansky’s zero-divisor conjectureMain theorem extends its counterexample to odd characteristic, using its Sections 3–5
2610.17411On the generators of separable unital C*-algebrasGeorge A. Elliott, Chun Guang Li and Zhuang Niu · math.OA · 12 October 2026OpenAI’s release, family 296: The generator problem for finite factorsMain results take its single-generation theorem as external input
2610.16832Matching Complexity for Second-Order Stationary Point with Sample Reusean author of the workQian Chen, Wendao Wu, Haihan Zhang, Chenheng Zhang, Chunyuan Zheng, Cong Fang, Haoxuan Li and Zhouchen Lin · math.OC · 12 October 2026Second-Order Stationarity with Common Random Losses: Matching Tolerance Bounds (arXiv 2609.28238)Extends its techniques to reusable samples (shared authors)
2610.16394Dynamical generalizations of the two-point Chowla conjectureBiao Wang and Shaoyun Yi · math.NT · 12 October 2026OpenAI’s release, family 007: Ordinary two-point correlations and the corrected Elliott conjectureMain theorem generalizes its two-point Chowla theorem and uses its estimates in the proof
2610.16059Baire measurable functions and automorphisms of the truth-table degreesBjørn Kjos-Hanssen · math.LO · 12 October 2026OpenAI’s release, family 241: Rigidity of the Turing degreesMain theorem adapts its Turing-degree rigidity proof
2610.18907Ordinary and measurable colorings of Euclidean three-spaceMikhail Koshelev, Andrey Kupavskii and Vsevolod Voronov · math.CO · 11 October 2026OpenAI’s release, family 158: The Euclidean plane cannot be colored with five colorsMain theorem assumes its character-law rigidity theorem and adapts its transfer
2610.15809A note on the p-adic inhomogeneous Littlewood conjectureSam Chow and Owen Jones · math.NT · 11 October 2026OpenAI’s release, family 022: The weak inhomogeneous Duffin–Schaeffer conjectureMain theorem deduced from its weak inhomogeneous Duffin–Schaeffer theorem
2610.15674Coprime pairs with equal sum of divisors, assuming smooth shifted primesShisheng Li and Zhong Zhuang · math.NT · 11 October 2026OpenAI’s release, family 011: Prime-factor statistics of p-1Checks its argument for p+1; main conclusion conditional on it
2610.15592The Deligne–Drinfeld conjecture after OpenAIThomas Willwacher · math.QA · 11 October 2026OpenAI’s release, family 008: The Deligne–Drinfeld conjectureSimplified exposition of its proof
2610.15559The Goldbach exceptional set: an exponent 7/12 and conditional refinementJános Pintz and Ingo Althöfer · math.NT · 11 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisMain theorem's zero-free half-plane hypothesis supplied by its uniform 7/8 theorem
2610.15483On the OpenAI proof that Landau-Siegel zeros do not existWilliam D. Banks · math.NT · 11 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisExposition of its Landau–Siegel argument with explicit constants
2610.15424Every n points in the plane contain (m1 + o(1)) n points with no unit distanceShisheng Li and Yong Su · math.CO · 11 October 2026OpenAI’s release, family 158: The Euclidean plane cannot be colored with five colorsMain theorem's only non-classical input is its Haar rigidity theorem
2610.15363Polynomial Bohnenblust–Hille Bounds and Hardy–Littlewood Inequalities for Finite Local SystemsAnderson Barbosa, Daniel Núñez-Alarcón, Daniel Pellegrino and Anselmo Raposo · math.FA · 11 October 2026Polynomial growth of Bohnenblust–Hille constants on the Hamming cube (arXiv 2609.12427)Improves the polynomial bound on the Boolean Bohnenblust–Hille constants that it first proved to C·d^(5/2)(1 + log d)³, adapting its contraction argument
2610.15293A note on the logarithmic exponent of Ramsey number r(4,t)Kaizhe Chen · math.CO · 11 October 2026OpenAI’s release, family 170: Sharp logarithmic exponents for off-diagonal Ramsey numbersMain result deduced from its Ramsey and independence bounds
2610.15291Refining the 7/8 zero-free half-plane for Hecke and Dirichlet L-functionsHejian Sang · math.NT · 11 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisMain result improves its 7/8 half-plane, assuming its analytic inputs
2610.15066An Exact Hausdorff Measure as Natural Length for SLEDapeng Zhan · math.PR · 10 October 2026OpenAI’s release, family 230: Exact Hausdorff gauges for SLEMain theorem takes its exact Hausdorff gauge theorem as input
2610.14638The Cartan determinant conjecture for representation-finite algebras over perfect fieldsLiang Chen · math.RT · 10 October 2026The Cartan determinant conjecture for representation-finite algebras (arXiv 2608.23132)Extends its theorem (Cartan determinant one for representation-finite algebras of finite global dimension) from algebraically closed to perfect fields, taking its grading over the algebraic closure as input
2610.14127A lower bound for additive complements to the primesArtyom Radomskii · math.NT · 9 October 2026OpenAI: Improved Long Gaps Between PrimesB(X) ≫ log X log log X for every additive complement B of the primes, from the author's lemma adapting its construction
2610.12853On the Imaginary Part of ζ′/ζFarzad Aryan · math.NT · 9 October 2026More than two thirds of the zeta zeros are simple and on the critical line (arXiv 2608.13637)An exposition of its proof, beside a zero-side formula for the imaginary part of ζ′/ζ
2610.12803A conditional bound Λ ≤ 0.13963 for the de Bruijn–Newman constantCetin Hakimoglu-Brown · math.NT · 8 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisMain bound conditional on its 7/8 half-plane
2610.12681The square function estimate for the cone in four dimensionsFaruk Temur · math.CA · 8 October 2026OpenAI’s release, family 079: Local smoothing in three dimensionsMain estimate derived from its local-smoothing and propagation statements
2610.12622Removing Boundary Smoothness from the Hot-Spots Conclusion in the Simple-Eigenvalue CaseKrzysztof Burdzy · math.AP · 8 October 2026OpenAI’s release, family 369: The hot spots conjecture for simply connected planar domainsMain theorem is a reduction to its smooth-domain theorem, taken as input
2610.12619A cyclotomic gluing construction for lattice sphere packingsan author of the workNihar Gargava, Matthias Gröbner and Michael Simkin · math.NT · 8 October 2026Stochastically evolving ellipsoids with symmetries (arXiv 2606.05105)Extends its lower bound for the lattice sphere-packing density to every large dimension (a shared author)
2610.12234Slightly improved zero-free half-planes for the quasi-Riemann hypothesisBaiying Liu · math.NT · 8 October 2026OpenAI’s release, family 003: The quasi-Riemann hypothesisMain result improves its 7/8 half-plane by varying its parameters
2610.11449The Erdos similarity problem for null sequences with adjacent ratios bounded belowKan Jiang and Lifeng Xi · math.MG · 8 October 2026OpenAI’s release, family 084: The geometric case of the Erdős similarity conjectureMain construction adapts its finite-tree construction
2610.11143Mean dimension and radius of comparisonGeorge A. Elliott and Zhuang Niu · math.OA · 8 October 2026OpenAI’s release, family 302: Radius of comparison equals half the mean dimensionMain theorem extends its reverse bound; arguments based on the release preprint
2610.10762Long intervals of consecutive composite values of polynomialsArtyom Radomskii · math.NT · 7 October 2026OpenAI: Improved Long Gaps Between PrimesAdapts its short-translate construction to give c log X log log X consecutive composite values of an irreducible polynomial
2610.10502Notes on Kozma and Nitzan's proof of θ(pc) = 0Arthur Blanc-Renaudie · math.PR · 7 October 2026θ(p_c) = 0 for Bernoulli bond percolation on ℤ^d in all dimensions d ≥ 2, in Lean 4 / MathlibA shorter, human-written proof of the Kozma–Nitzan inequalities, which give θ(p_c) = 0, reconstructed from its proof
2610.10502Notes on Kozma and Nitzan's proof of θ(pc) = 0Arthur Blanc-Renaudie · math.PR · 7 October 2026Kozma–Nitzan Connection Inequalities for Bernoulli PercolationA shorter, human-written proof of the inequalities it explains and extends, leaving to it the deduction of further Kozma–Nitzan conjectures
2610.10475A Strong Dominability Criterion and an Oka Union TheoremYun-Heng Du, Bin Guo, Peng-Chao Wang and Song-Yan Xie · math.CV · 7 October 2026OpenAI’s release, family 042: Oka classification for minimal compact complex surfaces: Kodaira dimension zero and class VIIMain theorem extends its Theorem 3.1 to all dimensions using its constructions
2610.10451Combs and complex structures on the six-sphereJeff Viaclovsky · math.AG · 7 October 2026The (3,4,∞) modular family of 2-tori, completed at its three special points, is a complex structure on S⁶Modifies its structure to give complex structures on S⁶ of arbitrarily large local Kuranishi dimension
2610.10192Irrationality exponents of logarithms of positive rational numbersJingwen Liu, Kai Jiang and Pingwen Zhang · math.NT · 7 October 2026OpenAI’s release, family 017: The irrationality exponent of π is 2Main theorem proved by adapting the determinant method of its π paper
2610.10188A proof of the fractional Lane-Emden conjectureLinfeng Mei and Juncheng Wei · math.AP · 7 October 2026OpenAI’s release, family 370: The Lane–Emden and Hénon–Lane–Emden conjecturesMain theorem proved by adapting its Hénon–Lane–Emden method
2610.10152The centered logarithmic Brunn-Minkowski inequalityYude Liu, Qiang Sun and Ge Xiong · math.MG · 7 October 2026OpenAI’s release, family 091: Logarithmic and Lp Brunn–Minkowski inequalities and the B-conjectureMain theorem extends its symmetric logarithmic Brunn–Minkowski proof to centered bodies, starting from its argument
2610.09370On the algebraic and analytic ranks of the twin-prime elliptic curve y^2=x(x-2)(x-p)Ruihan Chen · math.NT · 7 October 2026OpenAI’s release, family 002: The full BSD formula from low Selmer corankMain theorem conditional on its low-corank BSD statement
2610.08965More than 83.9% of the zeros of the Riemann zeta function are distinct and more than 67.35% are simple and on the critical lineKristian Muri Knausgård · math.NT · 6 October 2026More than two thirds of the zeta zeros are simple and on the critical line (arXiv 2608.13637)Improves both of its proportions by the same energy method, with a Lean formalization
2610.08581Counterexamples to the continuous Aldous-Lyons conjectureTsachik Gelander · math.GR · 6 October 2026OpenAI, Ten advances, result 3: Non-sofic groupsMain theorems take the existence of a finitely generated non-sofic group, credited to it with Fournier-Facio's and Kun–Thom's work, as input, giving non-co-sofic invariant random subgroups
2610.08144Navier-Stokes lost in translation: Why Lean verification of AI autoformalisation does not guarantee correct natural language proofsAlexander Bastounis, Fabian Circelli and Anders C. Hansen · math.AP · 6 October 2026OpenAI: Finite time blowup for Navier–StokesCompares its natural-language proof with its Lean formalization and argues that they do not correspond, making no claim about the proof itself
2610.06712Finite-time blowup for 2D unforced IPMMimi Dai · math.AP · 5 October 2026OpenAI: Finite time blowup for the Euler equationUnforced porous-media blow-up on the two-torus, taking from it the control of each stage's earlier history by the pressure Hessian
2610.06544Finite-time blowup of the IPM equation from smooth initial data without forcingFan Zheng · math.AP · 5 October 2026Extending the Córdoba-Martínez-Zoroa IPM Blow-Up to Uniformly Space-Time Smooth Forcing (arXiv 2609.16470)Removes the forcing from its blow-up for the incompressible porous media equation
2610.06544Finite-time blowup of the IPM equation from smooth initial data without forcingFan Zheng · math.AP · 5 October 2026OpenAI: Finite time blowup for the Euler equationAdapts its construction to prove unforced blow-up for the incompressible porous media equation on the two-torus
2610.02509Finite-time blowup of the Euler equation without forcingFan Zheng · math.AP · 1 October 2026OpenAI: Finite time blowup for the Euler equationAn exposition of its construction of blow-up for the unforced Euler equations
2610.00916Towards Strongly Aperiodic Monotiles in Higher DimensionsDmitry Kamenetsky · math.MG · 1 October 2026A Strongly Aperiodic Monotile in Three Dimensions (arXiv 2609.19214)Extends its three-dimensional strongly aperiodic monotile to ℝᴺ
2609.38442Complex structures on S6Philip Engel · math.AG · 29 September 2026The (3,4,∞) modular family of 2-tori, completed at its three special points, is a complex structure on S⁶An exposition and a self-contained proof of its theorem
2609.38408Large gaps between Romanoff numbersArtyom Radomskii · math.NT · 29 September 2026OpenAI: Improved Long Gaps Between PrimesAdapts its weight construction and its short-translates proposition to find ≫ log log X consecutive integers up to X that are not of the form p + 2ⁿ
2609.35406Finite-Time Blowup for Navier-Stokes with Smooth Forcing, Part I: Construction of Self-Similar Solutions with Admissible Stress and Flat RemainderZhen Lei and Xiao Ren · math.AP · 28 September 2026OpenAI: Finite time blowup for Navier–StokesAn exposition of the profile construction of its manuscript
2609.33785A two-parameter family of complex structures on S6Jeff Viaclovsky · math.AG · 27 September 2026The (3,4,∞) modular family of 2-tori, completed at its three special points, is a complex structure on S⁶A two-parameter family of complex structures on S⁶ from another elliptic surface, not biholomorphic to its examples
2609.33043More than 83.69% of the zeros of the Riemann zeta function are distinctKristian Muri Knausgård · math.NT · 27 September 2026More than two thirds of the zeta zeros are simple and on the critical line (arXiv 2608.13637)Improves its proportion of distinct zeros, following its proof
2609.31794The Wang Transform Inequalityan author of the workYair Lavi · math.CO · 25 September 2026The Marcus-Minc Transform Inequality (arXiv 2609.29262)Takes its lower estimate and bounds as inputs (the same author)
2610.00250Oka Geometry and Deformations of Torus-Fibred Complex Six-SpheresYun-Heng Du, Bin Guo, Yuta Kusakabe and Song-Yan Xie · math.CV · 23 September 2026The (3,4,∞) modular family of 2-tori, completed at its three special points, is a complex structure on S⁶A locally miniversal family on its six-sphere, every member of which is Oka
2609.28253A tyro's approach to the tilted sieve: beyond the Erdős–Rankin boundTristan Freiberg · math.NT · 23 September 2026A Tilted Residue-Class Construction for Long Prime-Free IntervalsAn elementary exposition of its tilted sieve and of its covering of composite survivors, without Maynard sieve weights or hypergraph covering, giving one residue class modulo each prime p ≤ x that covers an interval of length ≫ x log x/((log₂ x) log₃ x) (the paper is itself on the list)
2609.26706Conditional Oka Complex Structures of the Six-SphereZhangchi Chen · math.CV · 22 September 2026The (3,4,∞) modular family of 2-tori, completed at its three special points, is a complex structure on S⁶Proves that its threefolds are Oka manifolds
2609.26041A sharp density bound for 5-connected graphs with no K7 =  minorCaibing Chang, Zijian Deng, Qinfei Tang and Caihong Yang · math.CO · 22 September 2026Every graph with no K7 =  minor is 6-colorable (arXiv 2609.17760)Proves its Conjecture 1.4 by adapting its reduction method
2609.21159Erdős-Sós via random cyclic orderingsBryce Frederickson · math.CO · 18 September 2026A Counting Proof for Erdős Problem 548A simplified variant of its proof in the language of random cyclic orderings, with the same antidirected-tree conjecture
2609.20803Regularity of asymptotically axisymmetric solutions to the 3D Navier-Stokes equations with analytic forcingPeter Constantin, Mihaela Ignatova and Vlad Vicol · math.AP · 17 September 2026OpenAI: Finite time blowup for Navier–StokesTakes two properties of its construction as hypotheses and proves regularity under a real-analytic force, so its force cannot be real analytic near the singular point (the paper is itself on the list)
2609.17795An elementary way to find a counterexample to the Jacobian ConjectureArno van den Essen · math.AG · 15 September 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')Finds it again by elementary means, up to linear changes of coordinates
2609.27795A conditional construction of a nonhyperlinear group and the centralizer problemAndreas Thom · math.GR · 14 September 2026OpenAI, Ten advances, result 3: Non-sofic groupsA matrix counterpart of the Kun–Thom form of its mechanism, giving a nonhyperlinear group conditionally on the centralizer problem for matrix ultraproducts
2609.15893A short proof of the Erdős–Sós ConjectureOliver Riordan and Alex Scott · math.CO · 14 September 2026A Counting Proof for Erdős Problem 548A simplified version of its proof, with the extremal graphs and the antidirected-tree conjecture of Addario-Berry, Havet, Linhares Sales, Reed and Thomassé
2609.15080The Köthe conjecture via point modulesB. Greenfeld, B. King and L. Vendramin · math.RA · 14 September 2026Disproof of the Köthe conjecture (Krempa's matrix form)A new, short counterexample to the Köthe conjecture by another route, the point-module approach, after recalling that its counterexample came first
2609.29666Compact Complex Threefolds Fibred by Two-Dimensional Complex ToriXiufan Yang · math.AG · 13 September 2026The (3,4,∞) modular family of 2-tori, completed at its three special points, is a complex structure on S⁶Carries out its (3,4,∞) construction for the monodromy types (4,6) and (3,6)
2609.14273Disorder transfer for the critical SK overlap lawYan Ru Pei · math.PR · 13 September 2026Overlap distribution of the critical Sherrington-Kirkpatrick model (arXiv 2608.08752)Transfers its Gaussian overlap limit to other disorder, assuming it
2609.13906Failure of Rockafellar's Sum Conjecture via an Explicit Non-(FPV) OperatorRadu Ioan Bot · math.FA · 12 September 2026Nonmaximal sums of maximally monotone operators under Rockafellar's constraint qualification (arXiv 2609.10487)Simplifies and explains its counterexample to Rockafellar's sum conjecture
2609.11541A Complex Structure on A Rational Homology Six-Sphere with Two-TorsionZichang Wang · math.DG · 10 September 2026The (3,4,∞) modular family of 2-tori, completed at its three special points, is a complex structure on S⁶Adapts its construction, in Engel's form, to another elliptic surface, giving a simply connected rational homology six-sphere with H₂ ≅ H₃ ≅ ℤ/2
2609.10987Erdős-Sós for digraphsDhruv Mubayi and Jacques Verstraete · math.CO · 10 September 2026A Counting Proof for Erdős Problem 548The analogue for Eulerian digraphs, by its method (the proofs found by GPT-6 Astra; the paper is itself on the list)
2609.10269Singular Forces on the Whole Space: Sobolev Density Thresholds and Energy ApproximationShaozhen Cao and Zhuoni Chi · math.AP · 9 September 2026OpenAI: Finite time blowup for Navier–StokesThe counterpart on ℝ³: forces causing breakdown are dense in time-integrated Sobolev topologies exactly below a sharp threshold
2609.10262Density of Forces Producing Navier–Stokes BlowupShaozhen Cao, Zhuoni Chi and Ping Nie · math.AP · 9 September 2026OpenAI: Finite time blowup for Navier–StokesInserts its blow-up solution into regular trajectories: smooth forces producing blow-up from rest are dense on the torus in time-integrated Hˢ exactly when s < 1/2
2609.10085Finite-time divergence of kinetic-energy fluctuations in a Schrödinger-induction flowWeishuo Liu · math.AP · 9 September 2026OpenAI: Finite time blowup for Navier–StokesMain theorem takes its forced blow-up solution as hypothesis and carries it into a constrained Schrödinger–induction model, where the kinetic-energy fluctuations diverge in finite time
2609.08012Kalai's Conjecture for Tight TreesDhruv Mubayi and Jacques Verstraete · math.CO · 7 September 2026A Counting Proof for Erdős Problem 548Kalai's conjecture for tight trees in uniform hypergraphs, by extending its counting method (the proof itself found by GPT-6 Astra)
2609.06290The entropy production is not always monotone for the Kac equationWeinan Wang · math.AP · 5 September 2026The entropy production is not always monotone for hard spheres or for Maxwell molecules (arXiv 2609.01753)Adapts its two-temperature construction to the Kac equation
2609.04591Balanced Support Calibrations for Moser's Worm Problem: Exact Certificate and Bound of Triangular CoverZhipeng Deng · math.MG · 4 September 2026The exact solution of Bellman's lost-in-a-forest problem for the golden gnomon (arXiv 2607.24483)A lower bound in Bellman's escape problem by its balanced-support method and its Appendix A
2609.03736The half-rate linear programming bound for binary codes is 1/2−1/πAndrew Salmon · math.CO · 3 September 2026OpenAI, Ten advances, result 1: High-dimensional sphere packingProves the binary-code analogue of its sphere-packing theorem, the half-rate point 1/2 − 1/π of Delsarte's bound, following its mass-concentration argument
2609.03736The half-rate linear programming bound for binary codes is 1/2−1/πAndrew Salmon · math.CO · 3 September 2026OpenAI, Ten advances, result 2: Binary and spherical codesIts upper bound adapts its spherical-code construction to binary codes, improving its value at the half-rate point to exactly 1/2 (the paper is itself on the list)
2609.02882A new proof that more than 2/3 of the zeros of the Riemann zeta function are simple and on the critical lineYouness Lamzouri · math.NT · 2 September 2026More than two thirds of the zeta zeros are simple and on the critical line (arXiv 2608.13637)A new proof by a different method
2608.29651Remarks on the Complex Structures on ℙ3 and S2×S4Ruiming Liang, Chenhan Liu and Yang Zhang · math.DG · 30 August 2026The (3,4,∞) modular family of 2-tori, completed at its three special points, is a complex structure on S⁶Assumes its complex structure on S⁶ and builds from it complex structures on ℙ³ and on S² × S⁴
2608.27341Divisors in projective bundles over the projective line whose complement is affine spaceRemy van Dobben de Bruyn · math.AG · 27 August 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')A criterion for complements of divisors in projective bundles over ℙ¹ to be affine space, which explains its construction without coordinates
2608.18651A Uniform Proof for the Small Davenport Constant of the Exponent-p Heisenberg GroupAndreas Volkmann · math.CO · 19 August 2026The small Davenport constant of the Heisenberg group of order 125 (arXiv 2607.14379)The Godara–Sarkar conjecture d(H_{p³}) = 3p − 3 for every odd prime p, generalizing its case p = 5
2608.17862Idealizing Useful Fictions in Omega Grounded Arithmetican author of the workBryan Ford · math.LO · 18 August 2026Computable Quantification in Reflective Grounded Arithmetic (arXiv 2607.25533)A sequel by the same author: the system extended by an ω-grounded rule
2608.16140Internalized Truth in Reflective Grounded Arithmetican author of the workBryan Ford · math.LO · 17 August 2026Computable Quantification in Reflective Grounded Arithmetic (arXiv 2607.25533)A sequel by the same author: an internal truth predicate for the system, in Isabelle/HOL
2608.15681An exact formula for Erdős' problem 1005Yanmohan Wang, Mingxu Xie and Ziyuan Zhao · math.NT · 16 August 2026Optimality of Wouter van Doorn's Upper Bound for the Mayer-Erdős Farey Problem (arXiv 2607.23302)Makes its asymptotic for Erdős problem 1005 exact, following its framework
2608.14115The alpha spectrum of K-polystable toric ℚ-Fano varietiesXian Wu · math.AG · 14 August 2026K-polystable toric Fano varieties with small alpha invariants (arXiv 2607.04005)Answers its question on α-invariants of toric Fano varieties, generalizing its construction
2608.13747The small Davenport constant of the Heisenberg group of order 343Andreas Volkmann · math.CO · 13 August 2026The small Davenport constant of the Heisenberg group of order 125 (arXiv 2607.14379)The case p = 7 of the Godara–Sarkar conjecture by its method for p = 5
2608.12543Collision-Generated Compression for Homogeneous Keller MapsThomas Prellberg · math.AG · 12 August 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')A 40-variable counterexample to Zhao's vanishing conjecture, from Thompson's 24-variable cubic-homogeneous reduction of it
2608.06222Nonsofic wreath products of residually finite groupsGabor Kun and Andreas Thom · math.GR · 6 August 2026OpenAI, Ten advances, result 3: Non-sofic groupsStrengthens its Proposition 2.3 to show that generalized wreath products and doubles over infranormal property (T) subgroups are non-sofic
2608.05392The weak Markus–Yamabe conjecture fails in dimension 14Álvaro Castañeda, Gerardo Honorato and Francisco Valenzuela-Henríquez · math.AG · 5 August 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')The weak Markus–Yamabe conjecture fails in dimension 14, by a chain realization theorem applied to it
2608.03992On Diophantine equations over the integer rings of quadratic fieldsZhi-Wei Sun · math.NT · 4 August 2026An AI Proof of 18-Variable Undecidability for Diophantine Equations over ℤ[i] (arXiv 2606.12776)Hilbert's tenth problem over the integers of every quadratic field in 16 unknowns (15 for real fields), below its 18 over ℤ[i], using its Lemma 5
2608.02863Orbits and Fields of Definition for Graded Keller MapsKyle Kistner and Tanush Shaska · math.AG · 3 August 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')Classifies graded Keller maps by their weights, with new families, and shows it is the unique graded Keller map in its degree box
2608.02537Multicolor Ramsey numbers of odd cycles are superexponentialRaphael Steiner · math.CO · 3 August 2026OpenAI, Ten advances, result 9: Multicolor Ramsey numbersModifies its construction to make the multicolour Ramsey numbers of every odd cycle superexponential
2608.02522The Erdős-Hajnal conjecture for odd-girthRaphael Steiner · math.CO · 3 August 2026OpenAI, Ten advances, result 9: Multicolor Ramsey numbersA single-exponential bound in the Erdős–Hajnal problem for odd girth, deduced from its superexponential lower bound for multicolour triangle Ramsey numbers
2608.02025A torsion-free non-sofic groupFrancesco Fournier-Facio · math.GR · 3 August 2026OpenAI, Ten advances, result 3: Non-sofic groupsA finitely presented torsion-free non-sofic group, by applying its soficity criterion (its Proposition 2.3) to a small-cancellation construction
2608.01040The equality case of Ehrhart's volume conjectureJihao Liu · math.CO · 2 August 2026OpenAI, Ten advances, result 8: Ehrhart’s volume conjectureProves the equality case of Ehrhart's volume conjecture, with the frame and lemmas of its proof of the inequality
2608.00222Counterexamples to the Jacobian conjecture in dimensions greater than twoShuhong Gao · math.AG · 31 July 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')A self-contained account of its tangent-sweep mechanism, generalized to direction fields on hypersurfaces, giving counterexamples in dimensions above two
2607.24724Exponentially Many Circuit Double CoversRadek Hušek and Robert Šámal · math.CO · 27 July 2026OpenAI: A Proof of the Cycle Double Cover ConjectureReformulates its proof and uses its system of linear equations to count circuit double covers
2609.20220The Serre–Hochschild plane of a finite-dimensional algebraan author of the workMarco Armenta · math.KT · 26 July 2026τ-Hochschild (co)homology, the square of the Serre bimodule, and the Coxeter automorphism of the Tamarkin–Tsygan calculus (arXiv 2607.10913)Takes its Theorem 3.2 as a main input of several results (the same author)
2607.22198A five-variable counterexample to the Hessian conjecture, and the low-dimensional status of the Jacobian and Hessian conjecturesGuowu Meng and Liang Yang · math.AC · 24 July 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')A five-variable counterexample to the Hessian conjecture, from its six-variable doubling by a one-variable Schur descent
2607.21572Generic degrees of real polynomial Keller maps with non-dense imagePiotr Migus · math.AG · 23 July 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')Every even generic degree at least 4 of real Keller maps with non-dense image, realized by a deformation family of it
2607.20968An Explicit Characteristic-2 Counterexample to the Separable Jacobian ConjectureIrit Huq-Kuruvilla · math.AG · 23 July 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')A counterexample to the separable Jacobian conjecture in characteristic 2, by a modification of it (it does not itself specialize)
2608.23777An Explicit Counterexample to the Rank-Two Poisson ConjectureChristopher D. Long · math.RA · 22 July 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')A counterexample to the rank-two Poisson conjecture, built from its determinant-one normalization
2607.20210Graded Keller maps and the Jacobian ConjectureT. Shaska · math.AG · 22 July 2026A counterexample to the Jacobian conjecture in dimension 3 (X post 'Hello there the Jacobian conjecture is false')Studies its grading, fibres and rational points, and shows that the sign pattern of a grading decides whether graded Keller maps can be counterexamples
2607.16356A proof of the cycle double cover conjecture by OpenAI: An expositionSang-il Oum · math.CO · 17 July 2026OpenAI: A Proof of the Cycle Double Cover ConjectureA self-contained exposition of its proof, with modifications
2607.15501Cycle lengths and chords under chromatic and degree constraintsXiaozheng Chen and Bo Ning · math.CO · 16 July 2026Short proofs in combinatorics, probability and number theory II (arXiv 2604.06609)Extends its counterexample to Voss's conjecture from k = 4 to every k ≥ 5
2607.15399OpenAI's proof of the Cycle Double Cover TheoremJim Geelen · math.CO · 16 July 2026OpenAI: A Proof of the Cycle Double Cover ConjectureA short exposition of its proof
2607.05374The Minkowski grid has robustly many repeated distancesSungchul Lee, Cosmin Pohoata and Daniel G. Zhu · math.CO · 6 July 2026OpenAI: Planar Point Sets with Many Unit DistancesA robust form of its counterexample by a construction of the same kind: every large subset of the grid has a much-repeated distance
2606.27352Congruent copies of finite patterns in planar point setsShubhrajit Bhattacharya and Ritesh Goenka · math.CO · 25 June 2026OpenAI: Planar Point Sets with Many Unit DistancesExtends its lower bound from unit distances to congruent copies of any finite planar pattern
2606.16738A counterexample to the near-quadratic Elekes–Rónyai expander conjecture over ℝJihao Liu · math.NT · 15 June 2026OpenAI: Planar Point Sets with Many Unit DistancesDisproves the near-quadratic Elekes–Rónyai expander conjecture over ℝ, with its tower of number fields (its Proposition 3.8) as the main input
2606.03419Optimizing Explicit Unit-Distance Lower-Bound CertificatesMichael T. M. Emmerich · math.OC · 2 June 2026OpenAI: Planar Point Sets with Many Unit DistancesOptimizes the parameters of its construction, in Sawin's explicit form, to certify the exponent 1.0152 for the number of unit distances
2607.09667Undecidability of the quasi-isomorphism problem for semi-free dg algebras via the cobar constructionManuel Rivera · math.RA · 23 May 2026Undecidability problems for semifree DG algebras (arXiv 2605.08122)A new proof of its undecidability theorem by a different method

What counts

A follow-up paper is a new arXiv preprint whose main result builds on a document whose authors declare both its proofs and its writing generated by an AI model.

AI-generated works. A document counts as AI-generated when its authors, or its publisher, declare that an AI model generated both its proofs and its text. The list holds 938 works:

  • OpenAI’s October 6 release (github.com/openai/math, 2026), which says its manuscripts were “produced by an internal OpenAI model”: 724 manuscripts in 373 result families (722 in the first release, two added on 26 October, one of them the refutation of the Exponential Time Hypothesis), each family counted as one work.
  • OpenAI’s other mathematics documents, each saying its proofs were produced by an OpenAI model: the disproof of Erdős’s unit distance conjecture (May 2026), the proof of the cycle double cover conjecture (July), “Ten advances in mathematics and theoretical computer science” (August; each of its ten results counted as one work), the improved upper bound for short gaps between primes (August), the improved lower bound for long gaps between primes, the finite-time blow-up for the Navier–Stokes and for the Euler equations (September), and the construction of quantum Yang–Mills theory whose Lean code OpenAI submitted in October to the Yang–Mills Shepherding Committee, which calls it an LLM-produced result.
  • 279 arXiv papers of January to October 2026, in the categories below, whose statement on AI use says a model generated the text and also says a model produced the proofs or the main results. Papers whose statement covers only the writing, credits the model with a named part only (a key idea, an example), or leaves the extent of its part vague are not on the list. Papers of November 2026 join the list once the month closes.
  • 257 works on Hexagon, deposited from September 2026 to 16 November 2026, whose depositor declares the text primarily AI-generated and whose provenance statement says a model produced the proofs, by the same reading as for arXiv papers.
  • Twelve documents added by hand: the manuscript constructing a complex structure on S6 that Claude produced under the direction of L. Alpöge; the proof of the Erdős–Sós conjecture (Erdős problem 548) found by GPT-6 Astra, with the write-up generated from it; the argument improving the lower bound for gaps between primes (Erdős problem 4) posted on erdosproblems.com in August 2026, whose PDF names GPT 5.6 Sol as its author (the site credits GPT 5.6 Pro); and Anthropic’s refutation of the Exponential Time Hypothesis, “A 2o(n) Time Algorithm for 3-SAT” (October 2026), whose byline is Claude and whose core algorithm, it says, an internal Claude model found with no human input; and six results that Wikipedia lists among mathematical discoveries made with AI and whose own account says the model found them: four by GPT-6 Astra in Epoch AI’s evaluations (the disproofs of the Köthe and Ibragimov–Iosifescu conjectures, the proof of the Dittert conjecture, a counterexample to the γ–θ conjecture on eternal domination), Claude’s proof that θ(pc) = 0 for bond percolation on ℤd, and the Kozma–Nitzan connection inequalities produced by ChatGPT 5.6 Sol and Claude Fable 5.1 (September 2026); and the counterexample to the Jacobian conjecture in dimension 3 that L. Alpöge posted on X in July 2026, crediting the Claude model Fable “for working” on it; and the smooth counterexamples to the Loewner and Carathéodory conjectures that he posted in August 2026 after looking at the problem “with a model”, checked, he says, by himself, J.-P. Smith and Claude.

Follow-up papers. A paper follows up an AI-generated work when its main result, as its abstract and introduction present it, takes that work’s result as an input or a hypothesis; extends, improves or makes it explicit; adapts its method; or gives a new proof, an exposition, a formalization or a check of it. Papers that cite an AI-generated work for background, for comparison, to state priority, or in a remark are not counted; the table above gives their number. A paper sharing an author with the work it follows up is counted like any other.

The papers searched. Every new preprint from January 2026 to 13 November 2026 whose primary category is one of 28 arXiv mathematics categories: all of them except General Mathematics, History and Overview, Information Theory and Statistics Theory, with Mathematical Physics. Each paper is read from the source of its first version, and its title, abstract, comments and every source file are searched for the arXiv number, the Hexagon number, the title or the address of a work on the list. Each paper found was then read in full where it cites the work, and classified by Claude; every follow-up reading was checked against the paper’s text. Months are months of submission to August 2026, and of the arXiv number from September 2026.

Disclaimers

The counts record what papers say they do with an AI-generated work, and say nothing about whether any of the mathematics is right.

Not a judgement of correctness. A paper is listed for what it says it does with an AI-generated work: using it, extending it, explaining it, or checking it. Being listed says nothing about whether that paper or the AI-generated work is correct, and some follow-up papers were written precisely to check, or to correct, the work they follow.

Declarations are taken as stated. Whether a document is AI-generated is taken from what its authors or publisher declare; nothing has been verified. A document whose authors say nothing about AI is not on the list, however it was produced.

The counts are lower bounds. Only papers in the categories above are searched, only first versions are read, and a paper that refers to an AI-generated work without naming its arXiv number, title or address (“a recent AI proof of …”) is not found. November 2026 is incomplete, and its own AI-generated arXiv papers are not yet on the list.

Readings can be wrong. Each classification is a judgement made by reading the paper. Corrections are welcome; get in touch.

A study directed by Kimball Martin using Claude. Corpus: every new preprint in 28 arXiv mathematics categories from January 2026 to 13 November 2026, read from the arXiv source of its first version. A companion page follows AI use in arXiv mathematics papers. Corrections, references and discussion of each manuscript of OpenAI’s October 6 release are collected by an independent community effort, oai-math-revised. Page generated 17 November 2026.