Skip to content

Erdős Problem #143

Editorial depth: Reviewed explanatory article — evidence qualified.
Snapshot classification: open (not automatically the live result).
Subject areas: primitive sets.

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

Let \(A\) be a countably infinite subset of the real interval \((1,\infty)\). Assume for any distinct \(x,y\in A\) and every positive integer \(k\) that

\[ |kx-y|\ge1. \]

Does this multiplicative separation force the elements of \(A\) to be sparse? Erdős asked especially whether \(\sum_{x\in A}1/(x\log x)\) converges, and whether the restricted harmonic sum is \(o(\log N)\). These are distinct questions; sublogarithmic partial harmonic sums alone do not imply convergence of the extra logarithmically weighted series.

A small example

If \(A\) consists only of positive integers, the inequality forbids \(y=kx\) for any distinct \(x,y\in A\), making \(A\) a primitive set. This links the problem to classic work of Erdős and Behrend. The hypothesis permits noninteger real elements, which cannot simply be replaced by integers without proof. Since \(k=1\) is allowed, all distinct points of \(A\) are also separated by at least one in ordinary distance.

What is known

PARTIALLY SOLVED; the stronger convergence question remains OPEN. The maintained Erdős #143, updated April 2026, credits Koukoulopoulos, Lamzouri and Lichtman (2025) with the harmonic-sum conclusion

\[ \sum_{\substack{x<N\\x\in A}}\frac1x=o(\log N). \]

Its main question whether \(\sum_{x\in A}1/(x\log x)<\infty\) remains open. For integer primitive sets, Erdős established the stronger convergence statement, while other authors obtained counting estimates. The real-valued extension is material and not automatic.

GCL's current state

Pinned OPEN, original GCL issue #1412. The frozen Formal Conjectures 143 entry expresses the geometric separation and summability question, with proof holes; its historical sublogarithmic variant is a comment rather than a completed formal theorem. This difference is an editorial-fidelity boundary.

The original source assignment has an authenticated RESULT/1 report. This is unadjudicated source evidence, not a proof certificate or accepted article.

What remains

Prove the elementary consequence that the \(k=1\) hypothesis implies separation by at least one, then build exact finite predicates on rational sets and test the stronger \(k\ge1\) conditions for bounded multipliers. Acceptance: each test reports a bound on checked \(k\), does not conflate finite checking with the full quantified property, and distinguishes the open weighted-series convergence from the proven harmonic estimate.

Sources and claim boundaries

No proof of the logarithmically weighted convergence question or editorial admission is claimed.

Catalogue and evidence provenance

  • Pinned catalogue source: teorth/erdosproblems at 536408716724.
  • GCL issue: #1412 — protected issue/assignment identity; its labels and results are live, not reproduced automatically here.
  • GCL triage: PROTECTED_READY_BACKLOG.
  • 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.

Back to catalogue · Compendium introduction