Cumulants of threshold for Schensted row insertion into random tableaux

Abstract

Schensted row insertion is a fundamental component of the Robinson-Schensted-Knuth (RSK) algorithm, a powerful tool in combinatorics and representation theory. This study examines the insertion of a deterministic number into a random tableau of a specified shape, focusing on the relationship between the value of the inserted number and the position of the new box created by the Schensted row insertion. Specifically, for a given tableau and a point on its boundary, we consider the threshold that separates values which, if inserted, would result in the new box being created above the point from those that would result in a new box below. We analyze a random tableau of fixed shape and study the corresponding random threshold value. Explicit combinatorial formulas for the cumulants of this random variable are provided, expressed in terms of Kerov's transition measure of the diagram. These combinatorial formulas involve summing over non-crossing alternating trees. As a first application of these results, we demonstrate that for random Young tableaux of prescribed large shape, the rightmost entry in the first row converges in distribution to an explicit Gaussian distribution.