2026/07/30 by Jérémy Champagne
Mathematics · #math.NT #msc:11N25 #msc:11K06
12 pages
arxiv created 2026/07/30 · arxiv updated 2026/07/31
We answer a question of Bergelson and Richter about the probability of the relation gcd(n,\lfloorα1n\rfloor,\lfloorα2n2\rfloor,...,\lfloorαknk\rfloor)=1 for n∈ℕ when α1,...,αk are fixed irrational numbers. In particular, we avoid the use of exponential sums, as they are difficult to control when one of α1,...,αk admits very efficient rational approximations. Instead, we use an elementary method similar to that of Erdős and Lorentz, together with some general bounds on those n's which share a large prime divisor with \lfloorα1n\rfloor.