Skip to content

Erdős Problem #11

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

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

An integer is squarefree when no prime square divides it. Erdős Problem #11 asks whether every sufficiently large odd integer can be written

\[ n=s+2^e,\qquad s>0\text{ squarefree},\quad e\ge0. \]

An older, stronger formulation asks whether every odd integer greater than one has this form. It would be an error to treat the two quantified statements as equivalent merely because many small integers have been checked. The large-integer assertion is still open in the reviewed source record.

A small example

The odd integer \(11\) can be written as \(7+2^2\): the remainder 7 is squarefree. Exponent 1 instead gives \(11-2=9\), not squarefree because \(3^2\) divides 9. Since exponent zero is also allowed in the formal statement, \(11=10+2^0\) is another valid representation. Thus the smallest positive exponent and the smallest nonnegative exponent can differ.

A squarefreeness test checks whether any prime square divides the remainder. A finite search over exponents and odd integers can verify examples, but cannot establish a claim for all sufficiently large odd integers.

What is known

Computational work reported by the maintained problem registry includes a check by Hercher through \(2^{50}\), far beyond the smaller illustrative GCL finite replay. The large finite threshold does not turn the universal problem into a theorem.

Granville and Soundararajan, A binary additive problem of Erdős and the order of 2 mod \(p^2\), The Ramanujan Journal 2 (1998), 283–298, relate stronger representation hypotheses to the distribution of non-Wieferich primes. Their implications are conditional. In particular an implication about the older all-odd hypothesis must not be stated as a consequence of the weaker eventual question without further proof.

GCL's current state

The source audit #1280 identifies the old global versus current eventual quantifier difference, the exponent-zero distinction, and the frozen Lean statement's unfilled proof. It reports an independent integer-only replay for all 499,999 odd numbers from 3 through 999,999 using positive exponents, with exact witness hashes and histogram, plus a separate exponent-zero analysis. That is useful reproducibility evidence but not an asymptotic proof or an independent replay of Hercher's much larger computation.

What remains

An affirmative proof must handle every odd integer beyond a single fixed threshold, not merely a finite prefix or most odd integers. A refutation needs arbitrarily large odd exceptions. The source work also leaves a formalization decision: whether to encode the current eventual question or separately preserve the older stronger variant.

Sources and claim boundaries

Catalogue and evidence provenance

  • Pinned catalogue source: teorth/erdosproblems at 536408716724.
  • GCL issue: #1280 — 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