Erdős Problem #38
Editorial depth: Reviewed explanatory article — evidence qualified.
Snapshot classification: proved (not automatically the live result).
Subject areas: number theory.
Read this status correctly
The status above is reproduced from GCL's immutable external-catalogue snapshot. It is not a new claim of current mathematical resolution, a source-semantic review, an independent proof check, or GCL certification. A published proof and a GCL replay are separate events.
The question in plain language
A set \(B\) is an essential component if adding it to any set of intermediate Schnirelmann density increases that density. Erdős Problem #38 asks for a stronger property: can \(B\) fail to be an additive basis, yet for every set \(A\) of Schnirelmann density \(\alpha\) strictly between zero and one and every cutoff \(N\), some single shift \(b\in B\) increase the number of covered integers in \([1,N]\) by at least a fixed positive fraction \(f(\alpha)\)?
The maintained problem registry presently reports PROVED (LEAN), attributing a 2026 positive solution to GPT 5.5 Pro prompted by a contributor. That is an external provenance claim, not evidence of an independently reproduced GCL certification.
A small example
Take \(A=\{1,3,5,\ldots\}\), the positive odd integers. Its Schnirelmann density equals \(1/2\). Choosing the shift \(b=1\) makes
The combined set has density one, a gain of \(1/2\) in this example. This does not exhibit the required set \(B\), since a valid \(B\) must simultaneously handle every admissible \(A\) and every cutoff while itself not being an additive basis.
A simple bound valid for any such universal improvement is \(f(\alpha)\le1-\alpha\), because coverage cannot exceed all integers.
What is known
The registry's May 2026 update reports a sparse random construction of \(B\) with improvement on the scale \(\alpha(1-\alpha)^2\). Its provenance is disclosed as a model-assisted solution discussed in the public forum, with externally reported Lean verification. Earlier work by Linnik provided an essential component that was not an additive basis, but that older, weaker fact does not establish the strong one-shift requirement.
GCL's current state
Source audit #1307 records the exact universal quantifiers, checks the elementary odd-integer example and identifies the need to replay the actual external theorem, proof and axiom transcript. This article reports what the maintained mathematical registry says; it does not claim GCL reviewed or certified the 2026 construction.
What remains
For independent GCL verification, obtain the precise public proof and formal declaration, inspect its dependence on the chosen \(B\), the cutoff \(N\) and the density \(\alpha\), then replay it with protected theorem and axiom controls. This is separate from the registry's reported solved status.
Sources and claim boundaries
- Erdős Problems #38: May 2026 positive-solution provenance and reported Lean status.
- GCL source return #1307: exact claim limits and elementary example, no internal proof certification.
Catalogue and evidence provenance
- Pinned catalogue source: teorth/erdosproblems at
536408716724. - GCL issue: #1307 — protected issue/assignment identity; its labels and results are live, not reproduced automatically here.
- GCL triage: OUTSIDE_STRICT_OPEN_TRIAGE.
- Formal-statement coverage: A provider statement was included in the pinned snapshot. Presence does not mean the statement or proof has been checked.
- External mathematical status: consult current problem page and its dated bibliography.