Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
162 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

The distribution of the maximum of cubic character sums (2410.22305v1)

Published 29 Oct 2024 in math.NT

Abstract: For a primitive Dirichlet character $\chi\pmod q$ we let [M(\chi):= \frac{1}{\sqrt{q}}\max_{1\leq t \leq q} \Big|\sum_{n \leq t} \chi(n) \Big|.] In this paper, we investigate the distribution of $M(\chi)$, as $\chi$ ranges over primitive cubic characters $\chi\pmod q$ with $(q,3)=1$ and $q\leq Q$. Our first result gives an estimate for the proportion of such characters for which $M(\chi)>V$, in a uniform range of $V$, which is best possible under the assumption of the Generalized Riemann Hypothesis. In particular, we show that the distribution of large cubic character sums behaves very differently from those in the family of non-principal characters modulo a large prime, and the family of quadratic characters. We also investigate the location of the number $N_{\chi}$ where the maximum of $|\sum_{n\leq N} \chi(n)|$ is attained, and show the surprising result that for almost all primitive cubic characters $\chi\pmod q$ with $M(\chi)>V$, $N_{\chi}/q$ is very close to a reduced fraction with a large denominator of size $(\log V){1/2+o(1)}$. This contradicts the common belief that for an even character $\chi$, $N_{\chi}/q$ is located near a rational of small denominator and gives a striking difference with the case of even characters in the other two families mentioned above, for which $N_{\chi}/q\approx 1/3$ or $2/3$ for almost all even $\chi$. Furthermore, in the case of cubic characters, the works of Granville-Soundararajan, Goldmakher, and Lamzouri-Mangerel show that if $M(\chi)$ is large, then $\chi$ pretends to be $\xi(n)n{it}$ for some small $t$, where $\xi$ is an odd character of small conductor $m$. We show that for almost all such characters, we have $M(\chi)=m{-1/2+o(1)}\big|L(1+it, \chi\overline{\xi})\big|.$

Summary

We haven't generated a summary for this paper yet.