Skip to content

Erdős Problem #6

Editorial depth: Reviewed explanatory article — evidence qualified.
Snapshot classification: proved (not automatically the live result).
Subject areas: number theory, primes.

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

Write the primes in increasing order as \(p_1,p_2,\ldots\) and their successive gaps as \(d_n=p_{n+1}-p_n\). Erdős Problem #6, originating with Erdős and Turán, asked whether infinitely many starting positions satisfy

\[ d_n<d_{n+1}<d_{n+2}. \]

This is a question about three successive gaps between four consecutive primes. It is proved, not an outstanding Erdős conjecture: Banks, Freiberg and Turnage-Butterbaugh established the affirmative answer in 2015 and proved arbitrarily long increasing and decreasing runs.

A small example

Four consecutive primes are \(17,19,23,29\), with gaps \(2,4,6\). They provide one strictly increasing run, but no finite example proves that there are infinitely many.

Why can arbitrarily long runs be constructed? Suppose exponents \(e_0<e_1<\cdots<e_m\) are positive integers. For shifts \(x_j=2^{e_j}\), the successive differences satisfy

\[ x_{j+1}-x_j >\sum_{i<j}(x_{i+1}-x_i)\quad(j\ge1). \]

Indeed, the sum on the right telescopes to \(2^{e_j}-2^{e_0}\), while \(2^{e_{j+1}}-2^{e_j}\ge 2^{e_j}\). This is a finite elementary inequality, not the deep assertion that these shifts can become consecutive primes.

What is known

William Banks, Tristan Freiberg and Caroline Turnage-Butterbaugh, Consecutive primes in tuples, Acta Arithmetica 167 (2015), 261–266, Corollary 2 (printed pp. 262–263), prove that for every prescribed finite run length there are infinitely many strictly increasing prime-gap runs and infinitely many decreasing runs. Their Theorem 1 uses the Maynard–Tao machinery to arrange selected elements of suitable prime tuples as consecutive primes. The elementary power-of-two inequalities alone cannot replace that deep sieve input.

GCL's current state

The GCL source audit compares the published theorem with the frozen Formal Conjectures statement and checks an important index convention: the pinned Lean definition uses a zero-based prime enumerator. A parameter in a formal variant counts the number of gap comparisons, not the number of gaps; translation from the published theorem is therefore shifted by one. The provider proof bodies and GCL audit are not a newly replayed, kernel-certified proof of the 2015 publication.

What remains

Nothing remains to resolve the historical existence question; it was answered affirmatively in the literature. Separate formalization work may still be needed to state, import and kernel-check the published theorem with exact index conventions and its external sieve dependencies. Publication here is educational exposition, not new theorem certification.

Sources and claim boundaries

Catalogue and evidence provenance

  • Pinned catalogue source: teorth/erdosproblems at 536408716724.
  • GCL issue: #1275 — 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.

Back to catalogue · Compendium introduction