Wiki
Wiki

Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.

Updated

Problem 117

../

claims/: The 1 claim page of Problem 117, one per claimant's result; the problem's standing derives from them.


Statement. Let h(n)h(n) be minimal such that any group GG with the property that any subset of >n>n elements contains some x≠yx\neq y such that xy=yxxy=yx can be covered by at most h(n)h(n) many Abelian subgroups.

Estimate h(n)h(n) as well as possible.

Status. Open on the site (label OPEN; page last edited 23 January 2026). The site's proof-claims tab carries one full proof claim, submitted 18 August 2026 by Guillaume Lecomte and recorded as a pending claim on its claim page: that log⁡2h(n)=n/2+o(n)\log_2 h(n)=n/2+o(n), so h(n)1/n→2h(n)^{1/n}\to\sqrt{2}.

Source. erdosproblems.com/117, accessed 2026-09-04. Cite as: T. F. Bloom, Erdős Problem #117, https://www.erdosproblems.com/117.

References.

  • [Er97f] Erdős, Paul, Some unsolved problems. Combinatorics, geometry and probability (Cambridge, 1993) (1997), 1-10. library card, chapter at pp. 1-10.
  • [Py87] Pyber, L., The number of pairwise noncommuting elements and the index of the centre in a finite group. J. London Math. Soc. (2) (1987), 287-295.

Formalization. None recorded.

Current assessment

The site's formulation asks to estimate h(n)h(n), the least number such that every group in which any n+1n+1 elements include two distinct commuting ones is a union of at most h(n)h(n) Abelian subgroups. The known bounds are exponential: Pyber [Py87] proved c1n<h(n)<c2nc_1^n<h(n)<c_2^n for absolute constants c2>c1>1c_2>c_1>1, the upper bound through the theorem that a finite group with at most nn pairwise non-commuting elements has center of index at most cnc^n, and the lower bound, which Erdős [Er97f] attributes to Isaacs, from extraspecial 22-groups; the library card is Pyber 1987. The base of the exponential was left open there.

One pending full claim, Lecomte 2026, asserts the sharp rate log⁡2h(n)=n/2+O(n(log⁡(n+2))3)\log_2 h(n)=n/2+O(\sqrt{n}(\log(n+2))^3), with the lower bound from extraspecial 22-groups and the upper bound from a pp-group analysis through alternating forms, Sylow decomposition and a reduction to the centralizer of the derived subgroup. It is unreviewed and unpublished, so the derived standing is claimed, with claim value answered since the problem asks for an estimate. Should the claim be accepted, the question would be answered at the level of the exponential rate; the error term would remain a further estimate. The search behind this account covered the site's page, its discussion and proof-claims threads, Zenodo and arXiv on 2026-10-07.

Linked library material

These entries are derived from explicit links on library pages. They are navigation only and do not by themselves record mathematical progress.