Table of Contents
Fetching ...

Fourier and small ball estimates for word maps on unitary groups

Nir Avni, Itay Glazer, Michael Larsen

Abstract

To a non-trivial word $w(x_{1},...,x_{r})$ in a free group $F_{r}$ on $r$ elements and a group $G$, one can associate the word map $w_{G}:G^{r}\rightarrow G$ that takes an $r$-tuple $(g_{1},...,g_{r})$ in $G^{r}$ to $w(g_{1},...,g_{r})$. If $G$ is compact, we further associate the word measure $τ_{w,G}$, defined as the distribution of $w_{G}(\mathsf{X}_{1},...,\mathsf{X}_{r})$, where $\mathsf{X}_{1},...,\mathsf{X}_{r}$ are independent and Haar-random elements in $G$. In this paper we study word maps and word measures on the family of special unitary groups $\left\{ \mathrm{SU}_{n}\right\} _{n\geq2}$. Our first result is a small ball estimate for $w_{\mathrm{SU}_{n}}$. We show that for every $w\in F_{r}\smallsetminus\left\{ 1\right\} $ there are $ε(w),δ(w)>0$ such that if $B\subseteq\mathrm{SU}_{n}$ is a ball of radius at most $δ(w)\mathrm{diam}(\mathrm{SU}_{n})$ in the Hilbert-Schmidt metric, then $τ_{w,\mathrm{SU}_{n}}(B)\leq(μ_{\mathrm{SU}_{n}}(B))^{ε(w)}$, where $μ_{\mathrm{SU}_{n}}$ is the Haar probability measure. Our second main result is about the random walks generated by $τ_{w,\mathrm{SU}_{n}}$. We provide exponential upper bounds on the large Fourier coefficients of $τ_{w,\mathrm{SU}_{n}}$, and as a consequence we show there exists $t(w)\in\mathbb{N}$, such that $τ_{w,\mathrm{SU}_{n}}^{*t}$ has bounded density for every $t\geq t(w)$ and every $n\geq2$, answering a conjecture by the first two authors. As a key step in the proof, we establish, for every large irreducible character $ρ$ of $\mathrm{SU}_{n}$, an exponential upper bound of the form $\left|ρ(g)\right|<ρ(1)^{1-ε}$, for elements $g$ in $\mathrm{SU}_{n}$ whose eigenvalues are sufficiently spread out on the unit circle in $\mathbb{C^{\times}}$.

Fourier and small ball estimates for word maps on unitary groups

Abstract

To a non-trivial word in a free group on elements and a group , one can associate the word map that takes an -tuple in to . If is compact, we further associate the word measure , defined as the distribution of , where are independent and Haar-random elements in . In this paper we study word maps and word measures on the family of special unitary groups . Our first result is a small ball estimate for . We show that for every there are such that if is a ball of radius at most in the Hilbert-Schmidt metric, then , where is the Haar probability measure. Our second main result is about the random walks generated by . We provide exponential upper bounds on the large Fourier coefficients of , and as a consequence we show there exists , such that has bounded density for every and every , answering a conjecture by the first two authors. As a key step in the proof, we establish, for every large irreducible character of , an exponential upper bound of the form , for elements in whose eigenvalues are sufficiently spread out on the unit circle in .
Paper Structure (19 sections, 54 theorems, 188 equations)

This paper contains 19 sections, 54 theorems, 188 equations.

Key Result

Theorem 1.1

For every non-trivial word $w\in F_{r}$ there exist $\delta(w),\epsilon(w)>0$ such that for every $n$, every $0<\delta<\delta(w)$, and every $g\in\mathrm{SU}_{n}$, If $n>8\ell(w)$, one can take $\epsilon(w)=\frac{1}{256(\ell(w)+1)}$. Otherwise, $\epsilon(w)=\frac{1}{2\cdot10^{4}\ell(w)^{3}}$ suffices.

Theorems & Definitions (103)

  • Theorem 1.1
  • Remark 1.2
  • Corollary 1.3
  • Theorem 1.4
  • Remark 1.5
  • Theorem 1.6
  • Theorem 1.7: Fourier bounds for high-degree characters
  • Theorem 1.8: Fourier bounds for low-degree characters
  • Theorem 1.9: Fourier bounds for power words
  • Remark 1.10
  • ...and 93 more