← Back to arXiv
arXivNumber TheoryarXiv:2610.28685

An Asymptotic for Non-Wieferich Numbers

The paper deals with two special families of positive integers and shows that they almost completely overlap. The first family is called cyclic numbers, which are integers that have a certain property related to how many numbers smaller than them share no common factors with them. These were studied by the famous mathematician Paul Erdos, who figured out roughly how many cyclic numbers exist up to any large value x. The second family is called non-Wieferich numbers, a concept rooted in a classical unsolved problem in number theory about powers of 2. A Wieferich number is an odd integer that satisfies a specific divisibility condition involving the smallest power of 2 that is congruent to 1 modulo that integer. Non-Wieferich numbers are simply odd integers that fail this condition.

The central result of the paper is that the set of cyclic numbers and the set of non-Wieferich numbers are, in a precise asymptotic sense, almost identical. More specifically, the authors count how many integers up to x belong to one of these sets but not the other, which is called the symmetric difference. They prove that this count grows much more slowly than the count of either set individually, meaning the two sets diverge on only a negligible fraction of integers. This result does rely on a mild unproven assumption about a related concept called Wieferich primes, which are extremely rare primes satisfying a special congruence condition involving powers of 2.

The significance of this finding is that it connects two seemingly unrelated number-theoretic concepts and reveals a deep structural similarity between them. Cyclic numbers are defined through the arithmetic of Euler's totient function, while non-Wieferich numbers come from the order of 2 in modular arithmetic, so their near-equivalence is surprising and illuminating. The paper builds on work by Franco and Pomerance, who showed that non-Wieferich numbers are sparse among odd integers, and extends that line of research by giving a much more refined quantitative picture of how these two sparse sets relate to each other.

Read original →