Skip to content

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

OAI-26 family 013: Ostmann's inverse Goldbach conjecture

Editorial state: SOURCE_CRITIC_COMPLETE; protected acceptance is determined by the admissions ledger. Mathematical state: NO_CERTIFICATION. This article explains the pinned source claim, not an independently verified proof.

The question in plain language

The classical Goldbach problem concerns expressing even integers as sums of primes. An inverse question instead asks whether the collection of primes itself could be built as a sum of two simpler infinite sets, apart from finitely many exceptions.

For two sets of nonnegative integers \(A\) and \(B\), their sumset consists of all numbers \(a+b\) formed with \(a\) in \(A\) and \(b\) in \(B\). A decomposition is called nontrivial when both sets contain at least two elements. Ostmann's inverse Goldbach conjecture predicts that no set differing from the primes in only finitely many places can have such a decomposition.

The pinned September 24, 2026 manuscript The additive indecomposability of the primes asserts this as Theorem 1.1. This remains a source-attributed claim, not an independently checked GCL theorem.

A small example

If \(A=\{0,2\}\) and \(B=\{1,4\}\), then the sumset is \(A+B=\{1,3,4,6\}\). Its structure forces several pairs of numbers to appear together. That rigidity suggests why prime numbers might resist a sumset decomposition: shifting every member of one component by two different values in the other imposes simultaneous additive constraints.

But a finite example shows neither the behavior of all primes nor the possibility of allowing finitely many changes. The full conjecture must defeat every pair of subsets of the nonnegative integers with at least two elements apiece, including infinite sets.

What is known

The locked OAI-26 publisher overview identifies Family 013 as Ostmann's inverse Goldbach conjecture and states:

No finite modification of the primes can be written as the sum of two subsets of the nonnegative integers, each containing at least two elements.

The overview attributes this proposed result to the manuscript The additive indecomposability of the primes dated September 24, 2026. The formulation is stronger than simply saying that the primes themselves are not such a sum: it excludes adding or removing any finite collection of integers as a repair.

The qualification “finite modification” is important. It allows disagreement at finitely many small integers but requires agreement at all sufficiently large integers. The original PDF, Theorem 1.1, states that if \(A,B\) are subsets of the nonnegative integers with at least two elements each, then the symmetric difference between \(A+B\) and the set of positive primes is infinite. This is equivalent to the finite-modification assertion. It has no additional hypotheses that \(A\) and \(B\) be infinite: Lemma 2.2 first excludes finite summands, while Theorem 2.3 treats the remaining two-infinite-summand case. The argument requires both eventual coverage of all primes and eventual exclusion of composite sums; a density bound alone does not suffice.

GCL verification state

The primary source is indexed at the immutable repository snapshot openai/math@1581d4cbdd69a0faf34b2089614fe2a814316a6, with exactly one Family 013 PDF. The pinned overview.tex confirms the title and overview claim. The primary 80-page PDF was independently retrieved and verified against Git blob e9a0bb4bdbf86c71a968416d19c5c20d739d504f (764,239 bytes; SHA-256 2c4ec07457be59763b37d2197852341a7e6615e419982f9f3705c5f3a94c2c3f). Its abstract, Theorem 1.1, Lemma 2.2, Theorem 2.3 and opening proof outline were checked.

Source-critical limit: The exact theorem statement and the finite-versus-infinite-summand reduction were checked against the primary PDF. The proof itself, including the sieve estimates, character decorrelation and final tree-statistic contradiction, has not been independently verified. The source establishes what is claimed, not that the claim is true.

No source-bound Lean theorem, proof replay, Comparator result, independent proof-level mathematical critic or MATHCERT certificate is established here. A documentary description must not be interpreted as mathematical certification.

What remains

The bounded primary-source critic has checked the exact theorem statement and the reduction from finite to infinite summands. The next substantive mathematical work is to verify the large-sieve estimate for finitely many shifts, quadratic and higher-order character bounds, the finite-field tree comparison, and the positive-statistic contradiction in the original 80-page proof.

For documentary admission, the exact revised article hash, source-critical statement review, mathematics Markdown test, compendium regression tests, generated HTML, and protected merge/readback must be recorded. These checks do not mathematically certify the claimed resolution of Ostmann's conjecture.

Sources and claim boundaries

Boundary: NO_CERTIFICATION. This account attributes the claim to its publisher and does not endorse proof correctness.


Indexed primary manuscript paths: