The overlap distribution at two temperatures for the branching Brownian motion

TL;DR

This paper analyzes the overlap distribution at two temperatures in branching Brownian motion, demonstrating convergence and comparing mean overlaps to the random energy model using novel decoration point process techniques.

Contribution

It introduces a new application of the decoration point process to study overlap distributions at different temperatures in branching Brownian motion.

Findings

Overlap distribution converges under extended extremal process

Mean overlap at different temperatures is smaller than in Derrida's model

First use of decoration point process in this context

Abstract

We study the overlap distribution of two particles chosen under the Gibbs measure at two temperatures for the branching Brownian motion. We first prove the convergence of the overlap distribution using the extended convergence of the extremal process obtained by Bovier and Hartung. We then prove that the mean overlap of two points chosen at different temperatures is strictly smaller than in Derrida's random energy model. The proof of this last result is achieved with the description of the decoration point process obtained by A\"id\'ekon, Berestycki, Brunet and Shi. To our knowledge, it is the first time that this description is being used.