Skip to content

Erdős Problem #1096

Editorial depth: Authored explanatory DRAFT — NOT editorially accepted.
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.

UNACCEPTED DRAFT — mathematical editorial review pending

This material is a provisional, source-qualified manuscript published for line-by-line scrutiny and real HTML/LaTeX testing. It has NOT received separate GCL editorial acceptance or independent mathematical certification. Its presence on this website is not evidence of a solution, reviewed coverage, or final publication.

Review the underlying draft on GitHub.

The question in plain language

Choose a real number \(q>1\) and form every finite sum of distinct nonnegative powers of \(q\):

\[ \sum_{i\in S}q^i,\qquad S\subset\mathbb Z_{\ge0}\ \text{finite}. \]

Arrange the distinct sums in increasing order \(0=x_1<x_2<\cdots\). Erdős Problem #1096 asks whether, when \(q\) is sufficiently close to \(1\) from above, the consecutive gaps approach zero:

\[ \lim_{j\to\infty}(x_{j+1}-x_j)=0. \]

This means all sufficiently late gaps are small, not merely that some infinite subsequence of gaps approaches zero.

A small example

For the exact rational value \(q=11/10\), consider only powers \(q^0,\ldots,q^8\). Their \(2^9=512\) subsets produce 512 distinct sums, with a minimum consecutive gap of \(46729/100000000\) and maximum gap \(1\) within this finite collection.

Adding more allowed powers introduces additional sums and can split gaps, but a finite set cannot establish the behavior of the entire infinite increasing sequence. The finite maximum of one in this truncated example is not a counterexample to convergence of full-sequence gaps.

What is known

The original sufficiently-small‑\(q\) question was proved. The maintained catalogue credits Erdős and Komornik (1998) with convergence of the gaps to zero for

\[ 1<q<\sqrt{q_1}\approx1.175, \]

where \(q_1\) is the second smallest Pisot–Vijayaraghavan number.

The distinction between a limit and a lower limit is important. The catalogue reports Feng's 2016 criterion that, for \(1<q<2\), the liminf of the consecutive gaps is zero exactly when \(q\) is not a Pisot number. Under additional hypotheses involving \(q^2\), Feng also proves convergence of all consecutive gaps to zero over a larger interval.

A Pisot number is an algebraic integer greater than one whose other algebraic conjugates lie strictly inside the complex unit disk. Certain Pisot parameters obstruct overly broad extensions of the original theorem.

GCL's current state

Authenticated source audit #2367 distinguishes convergence from liminf behavior and records the published parameter restrictions. It also computes a bounded rational example without mistaking it for a proof of the infinite property.

The pinned Lean statement is categorized as solved, but its theorem body has an unproved placeholder. The published analytic proofs have not been independently replayed by GCL.

What remains

Audit the Erdős–Joó–Komornik, Erdős–Komornik and Feng primary papers for the exact interval endpoints, the \(q^2\) Pisot condition and the quantification over the full infinite spectrum. Keep finite subset-sum experiments separate from those theorems.

Sources and claim boundaries

Erdős Problem #1096; Erdős–Joó–Komornik (1990), Erdős–Komornik (1998), Feng (2016); authenticated GCL audit #2367. The original question is externally proved, not newly solved here; this article is unaccepted.

Catalogue and evidence provenance

  • Pinned catalogue source: teorth/erdosproblems at 536408716724.
  • GCL issue: #2367 — 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