On the critical value of void percolation in high dimensions
Patrick Charbonneau, Marcus Michelen
Abstract
In the random Lorentz gas, a point tracer moves through spherical obstacles of (reduced) density Φ centered on a Poisson point process in d-dimensional space, Rd, and becomes localized once the vacant space stops percolating, at Φ p(d). The dimensional scaling of this void percolation threshold has remained elusive. Prior bounds, conjectural ones included, ranged from Ω(1) to O(d d), leaving even the scaling undetermined. We here prove that Φ p(d)=Θ(d), and in particular that 12\,d Φ p(d) 5\,d for all sufficiently large d. The upper bound counts hierarchical families of escape waypoints called tree embeddings; the lower bound is via a weighted second-moment estimate on an explicit ensemble of ballistic escape routes. A linear scaling of Φ p(d) excludes the d d growth suggested by analogy with other systems in the random first-order transition universality class and is consistent with the numerical estimate Φ p/d 3.40(5). Because the mean-field dynamical transition is likewise linear, Φ d 2.40\,d, our result fixes the activated window [Φ d,Φ p] as a bounded---rather than diverging---correction to mean-field dynamical arrest in sufficiently high d.
Create a lesson
Related papers
Nonequilibrium Spin-Glass Order in a Long-Range Quantum Ising Glass and Its Sensitivity to Noise
Giuseppe D'Auria, Alessandro Silva
Cell-type-specific modularity in E/I networks shapes dynamical regimes and computational capabilities
Emmanouil Giannakakis, Victor Buendía, Oleg Vinogradov et al.
Finite-width shift of the edge of chaos in deep neural networks
Arn Caesar
Neural Networks with Exciton-Polariton Waveguides
Huawen Xu, Ruiqi Bao, Lingyu Tian et al.
Optical Ceiling on Polaron Mobility in Disordered Molecular Semiconductors: Lessons from Eumelanin
Alvan Hunt
Glauber dynamics of the Sherrington--Kirkpatrick spin-glass model: a dynamical cavity method
Grzegorz Szamel