The Oka property of complex K3 surfaces
From OpenAI Aftermath
The Oka property of complex K3 surfaces concerns the conjecture of Xie and Zhao, also a question in a survey of Forstnerič and Lárusson, that every complex K3 surface is an Oka manifold: holomorphic maps defined near a compact convex subset of Cm can be approximated there by entire maps, in every source dimension. A manuscript in OpenAI's public mathematics collection (github.com/openai/math), dated 23 September 2026, claims this for every complex K3 surface, including nonprojective ones, and derives the Oka property for every minimal compact complex surface of Kodaira dimension zero and a classification for class VII surfaces. The family has no Lean formalization. This page summarizes the published claims and lists publicly available follow-up work. Statements are the respective authors' claims; the publication of a paper, code or a formal proof is not presented here as independent verification.
Background
- As recounted in the introduction of the manuscript:
- Gromov's Oka principle for elliptic bundles and Forstnerič's characterization of Oka manifolds by the convex approximation property are the framework. Buzzard and Lu proved that elliptic and Kummer K3 surfaces are dominable by C2, and Forstnerič and Lárusson strengthened this to strong dominability for Kummer surfaces.
- Kamenova, Lu and Verbitsky proved that the Kobayashi pseudodistance of every K3 surface vanishes. Alarcón and Forstnerič established the one-dimensional Oka-1 property for Kummer and elliptic K3 surfaces, which does not give the full Oka property. Chen, Gounelas and Liedtke, and Chen and Gounelas, studied families of genus-one curves on projective K3 surfaces, which the manuscript uses. Xie and Zhao proved that smooth (2,2,2) hypersurfaces in (P1)3 are Oka, established dense sets of Oka K3 periods, and formulated the conjecture.
The original results
- Claim. Theorem 1.1 of Every complex K3 surface is Oka states that for every complex K3 surface, every compact convex set K in Cm, every holomorphic map from a neighbourhood of K and every ε > 0, there is an entire holomorphic map within ε on K. Corollary 1.2 gives, through every point and every nonzero tangent vector, a holomorphic immersion of C with dense image, Zariski dense for projective surfaces. The manuscript derives the Oka property for Enriques surfaces and hence for all minimal compact complex surfaces of Kodaira dimension zero; a separate classification states that a minimal class VII surface is Oka exactly when it is a Hopf or Enoki surface, using the global spherical shell theorem of a companion manuscript that is not an input to the K3 theorem.
- Method, as described: a local one-variable enlargement property with passive parameters, obtained from two transverse complete strips, implies the convex approximation property through polynomial-automorphism concentration and removal of polytope faces. On projective K3 surfaces the strips come from the genus-one curve families of Chen and Gounelas; the general case uses the period-theoretic framework of Xie and Zhao, with Verbitsky's corrected orbit alternatives, together with different geometric constructions and degeneration arguments.
- Source. The manuscript Every complex K3 surface is Oka at commit
1581d4c. - Formalization. The repository has no Lean scope note for this family.
Public follow-up work
| Date | Work | Relation | Public claim | Public status (checked 20 October 2026) |
|---|---|---|---|---|
| 2026-10-07 | S. An, Y.-H. Du, B. Guo, P.-C. Wang and S.-Y. Xie, A Local Approximation Characterization of Oka Manifolds, arXiv:2610.10475 | Uses and extends | Characterizes Oka manifolds by a local approximation property and derives applications: smooth projective rationally connected manifolds are Oka, removing an algebraic subset of codimension at least two preserves this, and compact Kähler manifolds with nonnegative holomorphic sectional curvature or nef tangent bundle are Oka. The paper states that the K3 conjecture was proved by OpenAI following the Xie–Zhao strategy and that this proof supplies the surface case it generalizes to targets of arbitrary dimension. | arXiv preprint; version 1 of 7 October 2026, version 2 of 12 October 2026. |
| 2026-10-16 | J. Jardine, Lattice24, referee review in JJardine919/openai-math-audit (archived at Zenodo 10.5281/zenodo.23422995) | Review (AI-assisted audit) | Reports no gap and states that the reviewer read every section. Its two minor notes: the class VII classification depends on the companion manuscript's global spherical shell theorem, which is an open conjecture whose companion proof was not checked, though the K3 theorem does not use it; and several uniformity claims under deformation of the finite atlas are argued only briefly, which the reviewer traced and found to hold but would want written out. | Part of an audit of the collection's manuscripts without Lean that the audit describes as complete. The reviews are published under CC BY 4.0. |
Notes
- The earlier results are due to the authors named in the Background section, as recounted by the manuscript.
- The list covers follow-ups found in public sources up to the date above; it is not a priority or completeness claim.
