A note on polynomial equidistribution and recurrence in finite characteristic

Ethan Ackelsberg, Vitaly Bergelson · 2026

Read A note on polynomial equidistribution and recurrence in finite characteristic by Ethan Ackelsberg, Vitaly Bergelson (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-22143.txt.

Introduction

Ackelsberg and Bergelson revisit polynomial equidistribution and recurrence over function fields. Their note fixes two small errors in a 2016 paper by Bergelson and Leibman. That paper proved a Weyl-type equidistribution theorem in finite characteristic. The authors set the fixes inside the wider field of number theory and additive combinatorics in function fields. They also prove new results on intersective polynomials in finite characteristic. Those results describe them through algebraic, combinatorial, and dynamic traits.

Catalog record

CREATOREthan Ackelsberg, Vitaly Bergelson
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE54569
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.22143

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A note on polynomial equidistribution and recurrence in finite characteristic Ethan Ackelsberg, Vitaly Bergelson Abstract . This paper addresses the topic of equidistribution and recurrence for polynomial sequences over function fields. The main focus is to note and correct two small errors in [ BL16 ] , contextualized within the broader developing literature on number theory and additive combinatorics in function fields. Connected with the resolution of these issues, we also prove new results characterizing intersective polynomials in finite characteristic in terms of various algebraic, combinatorial, and dynamical properties. 2020 Mathematics Subject Classification 11K36 (Primary) 11J71, 11T55, 05D10 (Secondary) 1. Introduction The goal of this paper is to clarify several facts related to uniform distribution of polynomial sequences and recurrence over function fields. This paper was spurred by recent advances that have revealed two subtle errors in [ BL16 ] : (1) A misstatement in the final sentence of [ BL16 , Theorem 0.3] , erroneously describing a condition on a polynomial g as being both necessary and sufficient for the polynomial sequence g(n) to be well distributed, when the condition is sufficient but not necessary. (The main content of the theorem expressed in the preceding sentences is unaffected by this error.) (2) An overly stringent condition used in describing intersective polynomials in the remark after [ BL16 , Theorem 9.5] that leads to clashes with other proposed definitions in the literature based on various equivalent conditions for intersectivity of integer polynomials. These issues do not affect other results from [ BL16 ] . Nevertheless, recent developments related to (1) and (2) have led to some confusion about the validity of other related results, and finding corrections in the literature has until now required careful reading of footnotes and remarks scattered between different papers (the issue (1) was raised in [ CGLLW25 ] and addressed in [ A26 , Remark 2.2] ; a description and resolution of issue (2) appears in [ AB25a , p. 3, footnote 3] and [ AB25b , Remark 4.15] ). For these reasons, we have written this note to pinpoint exactly where errors occur within the text of [ BL16 ] and to settle in a more definitive manner what the necessary corrections are. We also produce some new results on closely connected topics. For example, we show that the set of nonzero values of an intersective polynomial (when properly defined) in finite characteristic is a van der Corput set, resolving a conjecture from [ LLW23 ] . We elaborate and resolve each issue in turn, addressing (1) in Section 3 and (2) in Section 4 . In Section 4 , we have aimed to provide a rather comprehensive treatment of the topic of intersective polynomials in finite characteristic, and the final subsection ( 4.3 ) includes a proof of equivalences between several properties of algebraic, combinatorial, and dynamical nature that all characterize intersectivity of polynomials in finite characteristic. For ease of readability, we include in Section 2 a review of the notation and terminology introduced in [ BL16 ] . Included within Subsection 2.3 is a small correction to the utilization of derivational degree in [ BL16 ] . 2. Notation and terminology In this section, we collect notation and terminology from [ BL16 ] that will be needed in the sequel. 2.1. Basic algebraic objects in finite characteristic The goal of [ BL16 ] is to establish finite characteristic analogues of classical results of Weyl [ W16 ] about the distribution of polynomial sequences mod 1. One of the first steps in formulating analogies is the development of a glossary for translating between the two settings. In correspondence with the classical algebraic objects \mathbb{Z} (the ring of integers), \mathbb{Q} (the field of rational numbers), and \mathbb{R} (the field of real numbers), we write \mathcal{Z} for the ring \mathbb{F}[t] of polynomials over a finite field \mathbb{F} , \mathcal{Q} for the field of rational functions \mathbb{F}(t) , and \mathcal{R} for the field of formal Laurent series \mathbb{F}((t^{-1})) , which is the completion of \mathcal{Q} with respect to the metric induced by the absolute value \left|\frac{u}{v}\right|=|\mathbb{F}|^{\deg{u}-\deg{v}}. Finally, in analogy with the traditional \mathbb{T} used to denote the compact group \mathbb{R}/\mathbb{Z} , we write \mathcal{T} for the compact quotient group \mathcal{R}/\mathcal{Z} . Whereas Weyl’s polynomial equidistribution theorem concerns the distribution of values of a polynomial g(n)\in\mathbb{R}[n] taken mod 1 as n runs through the integers, the subject of [ BL16 ] is the distribution of values of a polynomial \mathcal{Z} -sequence g(n)\in\mathcal{R}[n] viewed mod \mathcal{Z} . (A \mathcal{Z} -sequence is a sequence indexed by \mathcal{Z} .) The integers \mathbb{Z} and the torus \mathbb{T} have a deep relationship expressed by their duality (in the sense of Pontryagin). The function e(x)=e^{2\pi ix} usefully displays the duality between \mathbb{Z} and \mathbb{T} via the pairing \langle n,x\rangle=e(nx) : this pairing is a group homomorphism in each variable, every character on \mathbb{Z} can be expressed uniquely as \langle\cdot,x\rangle for some x\in\mathbb{T} , and every character on \mathbb{T} can be expressed uniquely as \langle n,\cdot\rangle for some n\in\mathbb{Z} . The finite characteristic objects \mathcal{Z} and \mathcal{T} possess an analogous duality. For x=\sum_{i=-\infty}^{r}x_{i}t^{i}\in\mathcal{R} , we write e(x)=e^{2\pi i\textup{Tr}_{\mathbb{F}/\mathbb{F}_{p}}(x_{-1})/p}, where p is the characteristic of \mathbb{F} and \textup{Tr}_{\mathbb{F}/\mathbb{F}_{p}} is the trace map induced by viewing \mathbb{F} as a field extension of its prime subfield \mathbb{F}_{p} . (If |\mathbb{F}|=p^{k} , the trace can be written explicitly as \textup{Tr}_{\mathbb{F}/\mathbb{F}_{p}}(x)=x+x^{p}+\ldots+x^{p^{k-1}} .) Then \langle n,x\rangle=e(nx) represents characters on \mathcal{Z} and \mathcal{T} in its different variables in a manner directly analogous to the duality between \mathbb{Z} and \mathbb{T} described above. While the reader may at this point be distressed by the two distinct meanings of e(\cdot) presented in this paragraph, we assuage this valid concern by noting that e(\cdot) will from now on only be used to denote the function defined on \mathcal{R} in the finite characteristic setting. 2.2. Well distribution As mentioned above, we are interested in the distribution of values of polynomial sequences in tori \mathcal{T}^{c} . To discuss “distribution” in precise mathematical terms, we would like to determine the measures arising as limits of certain averages over terms of a sequence a(n)\in\mathcal{T}^{c} , n\in\mathcal{Z} . The relevant averaging method comes from the notion of a Følner sequence. A Følner sequence in \mathcal{Z} is a sequence of finite subsets (\Phi_{N})_{N\in\mathbb{N}} of \mathcal{Z} satifsying the asymptotic invariance property \lim_{N\to\infty}\frac{\left|(\Phi_{N}+n)\triangle\Phi_{N}\right|}{|\Phi_{N}|}=0 for every n\in\mathcal{Z} . For example, the sequence of sets \Phi_{N}=\{n\in\mathcal{Z}:\deg{n}\leq N\} forms a Følner sequence in \mathcal{Z} . We say that a \mathcal{Z} -sequence a(n)\in\mathcal{T}^{c} , n\in\mathcal{Z} , is well distributed with respect to a Borel probability measure \mu on \mathcal{T}^{c} if for every continuous function f:\mathcal{T}^{c}\to\mathbb{C} and every Følner sequence (\Phi_{N})_{N\in\mathbb{N}} in \mathcal{Z} , one has (2.1) \lim_{N\to\infty}\frac{1}{|\Phi_{N}|}\sum_{n\in\Phi_{N}}f(a(n))=\int_{\mathcal{T}^{c}}f~d\mu. The limiting measures that arise from polynomial sequences turn out to be supported on finitely many cosets of subgroups of \mathcal{T}^{c} , and we introduce specific terminology adapted to this situation. First, if X\subseteq\mathcal{T}^{c} is a closed subgroup and \mu=\mu_{X} is the Haar measure on X , we say that a sequence a(n) , n\in\mathcal{Z} , satisfying ( 2.1 ) is well distributed in X rather than directly referring to the measure \mu . More generally, if X=X_{0}+K is a finite union of cosets of a closed subgroup X_{0}\subseteq\mathcal{T}^{c} and

Read the complete text