Writing finite simple groups of Lie type as products of subset conjugates

TL;DR

This paper proves a near-optimal bound on expressing finite simple groups of Lie type as products of conjugates of a subset, improving previous results and using elementary combinatorial methods instead of heavy representation theory.

Contribution

It establishes a bound on the number of conjugates needed to express Lie type groups as products of a subset, approaching the conjecture with elementary techniques.

Findings

For any ε>0, G is a product of at most N_ε((log|G|)/log|A|)^{1+ε} conjugates of A or A^{-1}.

Improves bounds for symmetric sets compared to prior work.

Provides an alternative proof avoiding heavy representation theory machinery.

Abstract

The Liebeck-Nikolov-Shalev conjecture [LNS12] asserts that, for any finite simple non-abelian group $G$ and any set $A\subseteq G$ with $|A|\geq 2$, $G$ is the product of at most $N\frac{\log|G|}{\log|A|}$ conjugates of $A$, for some absolute constant $N$. For $G$ of Lie type, we prove that for any $\varepsilon>0$ there is some $N_{\varepsilon}$ for which $G$ is the product of at most $N_{\varepsilon}\left(\frac{\log|G|}{\log|A|}\right)^{1+\varepsilon}$ conjugates of either $A$ or $A^{-1}$. For symmetric sets, this improves on results of Liebeck, Nikolov, and Shalev [LNS12] and Gill, Pyber, Short, and Szab\'o [GPSS13]. During the preparation of this paper, the proof of the Liebeck-Nikolov-Shalev conjecture was completed by Lifshitz [Lif24]. Both papers use [GLPS24] as a starting point. Lifshitz's argument uses heavy machinery from representation theory to complete the conjecture, whereas this paper achieves a more modest result by rather elementary combinatorial arguments.