Nilpotent matrices having a given Jordan type as maximum commuting nilpotent orbit

TL;DR

This paper investigates the structure of nilpotent matrices with a fixed Jordan type, proving a table theorem and proposing a general box conjecture for partitions related to commuting nilpotent matrices.

Contribution

It proves the Table Theorem for partitions with two parts and introduces a generalized Box Conjecture for stable partitions, advancing understanding of nilpotent matrix classifications.

Findings

Proved the Table Theorem for specific stable partitions.

Formulated the Box Conjecture for arbitrary stable partitions.

Enhanced the classification framework for commuting nilpotent matrices.

Abstract

The Jordan type of a nilpotent matrix is the partition giving the sizes of its Jordan blocks. We study pairs of partitions $(P,Q)$, where $Q={\mathcal Q}(P)$ is the Jordan type of a generic nilpotent matrix A commuting with a nilpotent matrix B of Jordan type $ P$. T. Ko\v{s}ir and P. Oblak have shown that $Q$ has parts that differ pairwise by at least two. Such partitions, which are also known as "super distinct" or "Rogers-Ramanujan", are exactly those that are stable or "self-large" in the sense that ${\mathcal Q}(Q)=Q$. In 2012 P. Oblak formulated a conjecture concerning the cardinality of the set of partitions $P$ such that ${\mathcal Q}(P)$ is a given stable partition $ Q$ with two parts, and proved some special cases. R. Zhao refined this to posit that those partitions $P$ such that ${\mathcal Q}(P)= Q=(u,u-r)$ with $u>r\ge 2$ could be arranged in an $(r-1)$ by $(u-r)$ table ${\mathcal T}(Q)$ where the entry in the $k$-th row and $\ell$-th column has $k+\ell$ parts. We prove this Table Theorem, and then generalize the statement to propose a Box Conjecture for the set of partitions $P$ for which ${\mathcal Q}(P)=Q$, for an arbitrary stable partition $Q$.