Status
On this page
Status
Topics
Status
On this page
Status
Topics
Let be such that, for any set of size , the set
has size at least . Estimate .
Source: erdosproblems.com/539
No claim settles this problem.
The site's label is OPEN. The refereed bounds are : the lower bound is Erdős and Szemerédi's, by the pairing argument Granville and Roesler give on p. 2 of their paper, and the upper bound comes from the Freiman–Lev sets in two dimensions, which Granville and Roesler present (pp. 2--3) and record as their Theorem 2; the accepted partial claim Granville and Roesler 1999 records both, on the refereed publication alone. Erdős's 1973 announcement of the bounds (5.2) with Szemerédi has no claim page: it gives no proof and no reference, and its content is carried by the Granville–Roesler page, which credits it. Two further partial claims have pages. The exponent: Theorem A.1 of the arXiv preprint of July 2026 by Schmitt, Gehrunger, Dekoninck, Bérczi, Kreitner, Price and Holmes describing their system ProofCouncil, to which they attribute it, gives , hence ; the site's curator, Thomas Bloom, adopted the bound into the commentary on 15 June 2026 after sketching the construction himself, while the label stayed OPEN, and its claim page (Schmitt Gehrunger Dekoninck Berczi Kreitner Price Holmes, 2026) records it as a pending partial claim (not refereed; the adoption of a bound into the commentary of a problem the site labels OPEN is not an acceptance, and nothing is reviewed by this project). The lower bound: a Lean development of September 2026 states that , read as neither built nor audited by this corpus, on Kitamura's claim page (2026) (claimed). The authors of the exponent result write that the exact order of remains open within a subexponential factor, and this page reads the label OPEN the same way: the question asks for an estimate, and the order is not determined. No full claim exists, and the standing derives from the claim pages.