Wiki
Wiki

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

Updated


Claim. For every integer N≥1N\ge1, every Sidon set A⊆{1,…,N}A\subseteq\{1,\ldots,N\} and every ϵ>0\epsilon>0, there are arbitrarily large integers MM and a set B⊆{N+1,…,M}B\subseteq\{N+1,\ldots,M\} such that A∪BA\cup B is a Sidon set of size at least

(12−ϵ)M1/2.\left(\frac1{\sqrt2}-\epsilon\right)M^{1/2}.

Here a Sidon set is one whose sums a+ba+b over unordered pairs, repeated elements included, are pairwise distinct.

Covers. Problem 44 asks for size at least (1−ϵ)M1/2(1-\epsilon)M^{1/2} for every ϵ>0\epsilon>0. The claim gives the coefficient 1/21/\sqrt2 in place of 11, and so answers the question yes, for every NN and AA, at each ϵ>1−1/2\epsilon>1-1/\sqrt2. For 0<ϵ≤1−1/20<\epsilon\le1-1/\sqrt2, that is for the coefficient 11 itself, the question remains open.

Credit. The extension specializes Singer's prime-power difference-set theorem, consumed as an external premise at statement level; its proof is not reconstructed. Deleting finitely many conflicting differences from a Singer set and separating the three kinds of pair sums are Plasma AI's. No novelty is claimed.

Standing. L31 is author-recorded at tier 0, without independent examination or a Lean proof. Catalog acceptance is pending.

Depends on. L31 (tier 0), which consumes no other native claim; its one external premise is Singer's theorem, credited above.