🔢 mathematics

A Weyl equidistribution theorem over function fields

This paper establishes a new function field analogue of Weyl's equidistribution theorem for polynomials with irrational coefficients, thereby confirming a conjecture proposed by Lê, Liu, and Wooley.

Original authors: Ethan Ackelsberg

Published 2026-01-29

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ethan Ackelsberg

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: Spreading Out the Dots

Imagine you have a giant clock face, but instead of numbers 1 through 12, it's a continuous circle representing all the numbers between 0 and 1. This is what mathematicians call "mod 1."

Now, imagine you have a machine that spits out a sequence of numbers based on a formula (a polynomial). For example, the formula might be $P(n) = \text{something} \times n^2$ . As you feed the machine numbers $1, 2, 3, \dots$ , it spits out results. If you plot where these results land on the clock face, do they clump together in one spot, or do they spread out evenly across the entire circle?

The Goal: The paper proves that under certain conditions, these numbers will spread out perfectly evenly. No matter how small a slice of the clock face you pick, the machine will eventually land in that slice the exact right amount of times. Mathematicians call this "uniform distribution."

The Old Rule vs. The New Rule

The Classic Rule (Weyl's Theorem):
In the world of regular numbers (like 1, 2, 3.14, $\pi$ ), there is a famous rule from 1916. It says: If your formula has a "weird" number (an irrational number like $\pi$ or $\sqrt{2}$ ) in it, the results will always spread out evenly on the clock face.

The New World (Function Fields):
The author, Ethan Ackelsberg, is working in a different universe called Function Fields.

The Analogy: Instead of using regular numbers, imagine your numbers are made of polynomials (expressions like $t^2 + 3t + 1$ ).
The Twist: In this universe, there is a special "magic number" called $p$ (related to the size of the finite field, like a prime number).
The Problem: In this polynomial world, the old rule doesn't always work. Even if you have a "weird" coefficient, the results might not spread out evenly. They might get stuck in a corner.

Why does this happen?
The paper explains that the "magic number" $p$ acts like a trap. If your formula uses powers that are multiples of $p$ (like $n^p, n^{2p}, n^{3p}$ ), the math behaves like a photocopy machine that only prints perfect copies. It creates a rigid structure that prevents the numbers from spreading out.

The Main Discovery

The paper confirms a guess made by three other mathematicians (Lê, Liu, and Wooley). They guessed that the only reason numbers fail to spread out in this polynomial world is because of those "magic $p$ " powers.

The New Theorem (The "Safe Zone" Rule):
Ackelsberg proves that if you build your formula carefully, the numbers will spread out perfectly. Here are the three rules for a "safe" formula:

No Traps: At least one of the powers in your formula (like $n^k$ ) must not be divisible by the magic number $p$ . (e.g., $n^3$ is okay if $p=2$ , but $n^4$ is not).
No Multiples: None of the other powers in your formula can be a "multiple" of that safe power by a factor of $p$ . (You can't have $n^k$ and $n^{pk}$ in the same formula).
The Weird Ingredient: The number attached to that safe power ( $n^k$ ) must be "irrational" (in this specific polynomial sense, meaning it can't be written as a simple fraction of polynomials).

The Result: If you follow these three rules, your sequence of numbers will dance all over the clock face, visiting every spot equally often.

The "Photocopy" Analogy

Think of the polynomial sequence as a dance troupe.

Regular numbers: The dancers move randomly. If you give them a weird rhythm (irrational coefficient), they eventually cover the whole stage.
Function fields with $p$ -powers: The dancers are forced to move in a grid. If the rhythm is a multiple of $p$ , the choreography forces them to only step on specific tiles (like only stepping on red tiles). They never visit the blue tiles.
The Solution: The paper says, "If you remove the 'red tile' choreography (powers divisible by $p$ ) and ensure at least one dancer has a weird rhythm, they will break free from the grid and cover the whole stage."

What About More Variables?

The paper doesn't stop at one variable (like $n$ ). It also proves this works for formulas with many variables (like $n_1, n_2, n_3$ ). It's like having three dance troupes moving at once. As long as the "safe zone" rules are met for the combined movement, the whole group spreads out evenly.

Why This Matters (According to the Paper)

The paper is a "positive resolution" to a specific conjecture. It closes a gap in our understanding of how numbers behave in this abstract polynomial world. It tells us exactly when we can expect perfect randomness (equidistribution) and when we will get stuck in a pattern.

In short: The paper provides a recipe for creating perfectly random-looking sequences in a world of polynomial numbers, identifying that the only thing that ruins the randomness is a specific type of mathematical "locking mechanism" involving the number $p$ .

Technical Summary: A Weyl Equidistribution Theorem Over Function Fields

Problem Statement
The paper addresses the problem of uniform distribution (equidistribution) of polynomial sequences over function fields, specifically within the ring of polynomials $\mathbb{F}_q[t]$ and its field of rational functions $\mathbb{F}_q(t)$ . In the classical setting over the real numbers, Weyl's theorem (1916) establishes that a polynomial sequence $P(n)$ is uniformly distributed modulo 1 if at least one non-constant coefficient is irrational.

In the function field setting, where the base field $\mathbb{F}_q$ has characteristic $p > 0$ , the situation is more complex. Unlike the characteristic zero case, the presence of "irrational" coefficients (coefficients in the completion $\mathbb{R} = \mathbb{F}_q((t^{-1}))$ that are not in $\mathbb{Q} = \mathbb{F}_q(t)$ ) does not guarantee equidistribution. Specifically, if the exponents of the polynomial are divisible by $p$ , the Frobenius map $u \mapsto u^p$ acts as an additive homomorphism, creating obstructions to equidistribution. Furthermore, "interference" between exponents where one is a multiple of another by a power of $p$ can also prevent equidistribution.

Le, Liu, and Wooley ([LLW23]) conjectured that these specific structural obstructions (exponents divisible by $p$ and $p$ -power multiples) are the only barriers to equidistribution. This paper provides a positive resolution to their conjecture.

Methodology
The author employs a structural decomposition approach combined with existing ergodic-theoretic results. The methodology proceeds as follows:

Decomposition of Polynomials: Every polynomial $P(u) \in \mathbb{R}[u]$ is uniquely decomposed into a sum of additive polynomials applied to separable monomials. A monomial $u^r$ is "separable" if $p \nmid r$ . Additive polynomials are of the form $\eta(u) = \sum \alpha_i u^{p^i}$ . The decomposition takes the form:
$P(u) = \alpha_0 + \sum_{i=1}^d \eta_i(u^{r_i})$
where $\eta_i$ are additive polynomials and $u^{r_i}$ are distinct separable monomials.
Application of Bergelson-Leibman Theorem: The proof relies heavily on a general equidistribution theorem by Bergelson and Leibman ([BL16]). This theorem states that the closure of the image of a polynomial sequence is a union of cosets of a $\Phi$ -subtorus. Crucially, the sequence is well-distributed in the ambient torus $\mathbb{T}$ if the sum of the $\Phi$ -subtori associated with the additive components $\eta_i$ equals the entire torus $\mathbb{T}$ .
Analysis of Additive Components: The paper utilizes a specific result (Proposition 2.3, also from [BL16]) stating that for a linear additive polynomial $\eta(u) = \alpha u$ , the associated $\Phi$ -subtorus is the entire torus $\mathbb{T}$ if and only if $\alpha$ is irrational (i.e., $\alpha \notin \mathbb{Q}$ ).
Verification of Conditions: The proof demonstrates that under the conjectured conditions (existence of a term with exponent $k$ such that $p \nmid k$ , no $p^v k$ exists in the set of exponents, and the coefficient $\alpha_k$ is irrational), the decomposition of $P(u)$ yields an additive component $\eta_k(u) = \alpha_k u$ whose associated subtorus is $\mathbb{T}$ . This suffices to prove the well-distribution of the entire sequence.
Extension to Multivariable and General Fields: The argument is extended to multivariable polynomials and then to global function fields by utilizing integral bases to map the ring of integers of a global function field to $\mathbb{Z}^d$ , reducing the problem to the multivariable case over $\mathbb{F}_q[t]$ .

Key Contributions and Results
The primary result is Theorem 1.2, which confirms the conjecture of Le, Liu, and Wooley. It states:
Let $P(u) = \sum_{r \in K \cup \{0\}} \alpha_r u^r$ be a polynomial over the function field completion. If there exists an exponent $k \in K$ such that:

$p \nmid k$ (the exponent is separable),
For all $v \in \mathbb{N}$ , $p^v k \notin K$ (no $p$ -power multiples of $k$ exist in the exponent set),
$\alpha_k \notin \mathbb{Q}$ (the coefficient is irrational),
Then the sequence $(P(u))_{u \in \mathbb{Z}}$ is well-distributed in the compact group $\mathbb{T}$ .

The paper further provides:

Theorem 3.1: A multivariable extension of the result, establishing well-distribution for polynomials in $d$ variables under analogous conditions on the multi-indices and coefficients.
Corollary 3.3: An equidistribution theorem for polynomial sequences over the ring of integers of an arbitrary global function field, generalizing the result beyond $\mathbb{F}_q[t]$ .

Significance and Claims
The paper claims to provide a "positive resolution" to the conjecture regarding the necessary and sufficient conditions for Weyl-type equidistribution in function fields. By confirming that the only obstructions are those arising from the characteristic $p$ (specifically the Frobenius homomorphism and $p$ -power interference), the work clarifies the structural limitations of polynomial sequences in positive characteristic.

The author notes that while the proof of Theorem 1.2 appears in a separate work by CGLLW25 using different methods, this paper offers a distinct proof utilizing the Bergelson-Leibman framework. Additionally, the paper corrects a subtle error in the original formulation of the Bergelson-Leibman theorem regarding the necessity of the subtorus condition, clarifying that the condition is sufficient but not necessary for equidistribution in the general case, though it remains sufficient for the specific application in this paper.

The work is presented as a "short note" focused on resolving the specific conjecture, without proposing new experimental applications or future implications beyond the theoretical confirmation of the conjecture and its extension to multivariable and global settings.