Skip to content

Documentary state: EDITORIALLY_ADMITTED. The manuscript has a protected editorial admission, not a theorem certification.

OAI-26 family 059: Counterexamples to Zariski's multiplicity conjecture

Editorial status: DRAFT_UNACCEPTED pending protected admission. Mathematical status: NO_CERTIFICATION. This document checks the statements in pinned PDFs; it does not certify their proofs.

The question in plain language

The multiplicity of a hypersurface at a singular point is the degree of the lowest nonzero term of a local defining equation, when that equation is reduced. Zariski asked whether the shape of a hypersurface near a point, viewed up to a homeomorphism of the surrounding space, determines its multiplicity.

The two Family 059 manuscripts claim counterexamples: germs with different multiplicities but indistinguishable embedded topology.

A small example

Near the origin, an equation beginning in degree two has multiplicity two; an equation whose first nonzero terms have degree three has multiplicity three. Distinct multiplicities are easy to identify algebraically, but it is much harder to know whether their zero sets have the same topology inside the ambient complex space. Abstract homeomorphism of the zero sets alone would be weaker than the ambient homeomorphism required here.

What is known

The pinned September 24, 2026 paper Ambiently homeomorphic isolated hypersurfaces of multiplicities two and three states that it constructs reduced hypersurface germs with isolated singularities of multiplicities two and three in a common complex affine space of dimension divisible by eight. Its abstract also claims that the defining function germs are topologically right equivalent and that the examples bear on Arnold's corank question. The paper does not assert these examples lie in complex dimension four.

The pinned September 27, 2026 paper Ambiently homeomorphic isolated hypersurface germs in C4 with multiplicities four and five makes a separate four-variable claim in Theorem 1.1: reduced convergent holomorphic germs with isolated critical points, multiplicities four and five, and a homeomorphism of the corresponding ambient pairs. Its introduction describes using Milnor open books, integral Seifert forms and a classification result to compare their embedded links.

Taken as source claims, either construction would negate the embedded form of Zariski's multiplicity conjecture. The two pairs have different ambient-dimension constraints and should not be combined into a purported two-versus-three example in complex four-space.

GCL verification state

Source revision: openai/math@1581d4cbdd69a0faf34b2089614fe2a814316a6a. Both exact PDFs were downloaded, and their abstracts and opening theorem statements were inspected.

  • Two-versus-three manuscript: Git blob 0a388d1d0d16dd2cb75214dfba2cf834caf09e64, 560,543 bytes, 31 pages.
  • Four-versus-five manuscript: Git blob 8442e691b457ecfd824226dde58ec4e50955d7b0, 410,248 bytes, 32 pages.

This verifies the wording and scope of the manuscript claims, not the claimed constructions or their embedded topological equivalences. The Seifert-form computations, link classifications, isolation and reducedness checks, and exact equations have not been independently replayed. There is no source-to-Lean equivalence, Comparator replay or MATHCERT certificate.

What remains

A full proof-level review should reproduce the two sets of defining equations and verify their multiplicities, isolation properties, reducedness and ambient homeomorphisms. For the four-dimensional example, examine the integral Seifert-form congruence and invoked classification theorem. Separately test the higher-dimensional two-versus-three construction and its claimed topological right equivalence.

For documentary admission, retain a bounded source-critical record, run mathematical Markdown and compendium regression tests, inspect generated MkDocs HTML, register the exact article SHA-256 and merge through protected governance. Editorial acceptance is not mathematical certification.

Sources and claim boundaries

Boundary: NO_CERTIFICATION. Direct examination of theorem statements is not independent proof verification.


Indexed primary manuscript paths: