Status
On this page
Status
Topics
Status
On this page
Status
Topics
Prove that
for (except when ).
Source: erdosproblems.com/551
An accepted solution exists. The statement is true.
The site labels the problem DECIDABLE, which the site defines as resolved up to a finite check. The label is a statement about the shape of what remains, not a theorem; it is recorded as the accepted partial claim page Keevash, Long and Skokan 2021, whose Covers. states the finite check, and the page records what is proved and what remains. A full proof of the identity, closing that check, is given by the OpenAI release's preprint of 25 September 2026 and accepted on the claim page OpenAI 2026 on its Lean proof, which this corpus built, checked for axioms and found identical to the release's comparator challenge; the preprint itself is unrefereed and unreviewed. The derived standing, solved, proved, departs from the site's DECIDABLE by counting that accepted full claim; the refereed partial claims alone leave the finite check open, which is what the label records. Proved, in refereed sources, each an accepted partial claim on its refereed publication alone: the identity for ([BoEr73] Theorem 4, Bondy and Erdős 1973), for when ([Ni05], preprint Theorem 1, Nikiforov 2005) and for with an absolute constant that is not computed ([KLS21] Theorem 1.1), which covers every once exceeds a threshold that the paper does not name; the case is the classical for (quoted from Chartrand and Schuster on p. 47 of [BoEr73]), the cases , , are reported settled for all by the introductions of [Ni05] and [KLS21] (sources not held), and the literature list of [OAI26] (Section 1.1) reports settled for all by Chen, Cheng and Zhang (2008) and, for , the lengths , and settled by papers of 2007--2023 (sources not held). Left by the refereed results: for each of the finitely many with , the cycle lengths with , less the settled cases, so that for the lengths below the threshold of [KLS21] remain: a finite set of pairs whose extent is unknown because is not explicit. No refereed source closes it; the accepted 2026 result closes all of it. The one site comment (1 September 2025) describes the state before the release: reduced to a finitary problem, still open.