ProbXiv
Explorersign in
Problem archiveProblem record

Statement

Let U\mathbb U be the space of infinite upper-triangular unipotent matrices over the finite field with qq elements. For u∈Uu\in\mathbb U, let u(n)u^{(n)} be its top-left n×nn\times n corner, and let λi(u,n)\lambda_i(u,n) and λi′(u,n)\lambda'_i(u,n) denote the row and column lengths of the Young diagram parametrizing the Jordan normal form of u(n)u^{(n)}. A probability measure on U\mathbb U is central if its cylindrical probabilities depend only on conjugacy classes, and it is ergodic if it is an extreme point among central probability measures.

Classification and law of large numbers for ergodic central measures. Let ϑ\vartheta be an ergodic central measure on U\mathbb U. There exist nonincreasing sequences of nonnegative numbers r=(ri)r=(r_i) and c=(ci)c=(c_i) such that, for every ii, ϑ\vartheta-almost surely,

lim⁡n→∞λi(u,n)n=ri,lim⁡n→∞λi′(u,n)n=ci.\lim_{n\to\infty}\frac{\lambda_i(u,n)}{n}=r_i,\qquad \lim_{n\to\infty}\frac{\lambda'_i(u,n)}{n}=c_i.

Moreover, for every pair of sequences r=(r1≥r2≥⋯≥0)r=(r_1\ge r_2\ge\dots\ge0) and c=(c1≥c2≥⋯≥0)c=(c_1\ge c_2\ge\dots\ge0) satisfying ∑i(ri+ci)≤1\sum_i(r_i+c_i)\le1, there exists a unique ergodic central measure ϑr,c\vartheta^{r,c} with these row and column frequencies. Its cylindrical probabilities are given by

ϑ(Cyl⁡gGLU)=q−n(n−1)/2(1−q−1)nqn(λ)Spr,c(q),1[Qλ(⋅;q−1)],\vartheta(\operatorname{Cyl}^{\mathbb{GLU}}_g)=\frac{q^{-n(n-1)/2}}{(1-q^{-1})^n}q^{n(\lambda)}\mathcal{Sp}_{r,c^{(q)},1}\left[Q_\lambda(\mathord\cdot;q^{-1})\right],

where g∈GL(n,q)g\in\mathbb{GL}(n,q) is a unipotent upper-triangular matrix of conjugacy class λ\lambda, QλQ_\lambda is the Hall–Littlewood polynomial, and c(q)c^{(q)} is obtained by rearranging the numbers (1−q−1)ciq1−j(1-q^{-1})c_iq^{1-j}, for i,j=1,2,…i,j=1,2,\dots, in decreasing order. This conjecture proposes a complete classification of ergodic central measures and a law of large numbers for the shapes of their finite corners; its resolution would describe all such measures through their asymptotic row and column frequencies.

Record

Comments

Nothing has been published against this problem yet, and nothing has been said about it.

Sign in with an institutional address to take part in the discussion. Reading every thread stays open to everyone.

Sign in

Solve with an agent

Open the statement in a chat, with the problem and the ground rules already written into the prompt.

This opens a third-party site. Nothing is posted back to ProbXiv and nothing you write there is recorded here — what a model gives you is an attempt, which a person still has to check.