The frequency and the structure of large character sums
Abstract
Let M() denote the maximum of |Σn N(n)| for a given non-principal Dirichlet character q, and let N denote a point at which the maximum is attained. In this article we study the distribution of M()/q as one varies over characters q, where q is prime, and investigate the location of N. We show that the distribution of M()/q converges weakly to a universal distribution , uniformly throughout most of the possible range, and get (doubly exponential decay) estimates for 's tail. Almost all for which M() is large are odd characters that are 1-pretentious. Now, M() |Σn q/2(n)| = |2-(2)|π q |L(1,)|, and one knows how often the latter expression is large, which has been how earlier lower bounds on were mostly proved. We show, though, that for most with M() large, N is bounded away from q/2, and the value of M() is little bit larger than qπ |L(1,)|.
Turn this paper into a lesson
ArcXiv compiles a structured reading guide from this paper's metadata: plain-English importance, contributions, prerequisite concepts, which sections to read first, flashcards, and a quiz. Grounded in the abstract, never invented.