Large deviations of glassy effective potentials

TL;DR

This paper develops a theoretical framework for analyzing large deviations in glassy effective potentials, extending beyond near-transition regimes, with applications to spherical p-spin models and supercooled liquids.

Contribution

It introduces a general approach to study large deviations of disordered potentials in glassy systems, beyond the vicinity of the MCT transition.

Findings

Framework for large deviations in glassy potentials

Application to spherical p-spin models

Insights into memory effects in dynamics

Abstract

The theory of glassy fluctuations can be formulated in terms of disordered effective potentials. While the properties of the average potentials are well understood, the study of the fluctuations has been so far quite limited. Close to the MCT transition, fluctuations induced by the dynamical heterogeneities in supercooled liquids can be described by a cubic field theory in presence of a random field term. In this paper we set up the general problem of the large deviations going beyond the assumption of the vicinity to $T_{MCT}$ and analyze it in the paradigmatic case of spherical ($p$-spin) glass models. This tool can be applied to study the probability of the observation of a dynamics with memory of the initial condition in regimes where, typically, the correlation $C(t,0)$ decays to zero at long times, at finite $T$ and at $T=0$.