How to escape atypical regions in the symmetric binary perceptron: a journey through connected-solutions states

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autor principal: Barbier, Damien
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913769705177088
author Barbier, Damien
author_facet Barbier, Damien
contents We study the binary symmetric perceptron model, and in particular its atypical solutions. While the solution-space of this problem is dominated by isolated configurations, it is also solvable for a certain range of constraint density $α$ and threshold $κ$. We provide in this paper a statistical measure probing sequences of solutions, where two consecutive elements shares a strong overlap. After simplifications, we test its predictions by comparing it to Monte-Carlo simulations. We obtain good agreement and show that connected states with a Markovian correlation profile can fully decorrelate from their initialization only for $κ>κ_{\rm no-mem.\, state}$ ($κ_{\rm no-mem.\, state}\sim \sqrt{0.91\log(N)}$ for $α=0.5$ and $N$ being the dimension of the problem). For $κ<κ_{\rm no-mem.\, state}$, we show that decorrelated sequences still exist but have a non-trivial correlations profile. To study this regime we introduce an $Ansatz$ for the correlations that we label as the nested Markov chain.
format Preprint
id arxiv_https___arxiv_org_abs_2408_04479
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle How to escape atypical regions in the symmetric binary perceptron: a journey through connected-solutions states
Barbier, Damien
Disordered Systems and Neural Networks
Statistical Mechanics
Mathematical Physics
We study the binary symmetric perceptron model, and in particular its atypical solutions. While the solution-space of this problem is dominated by isolated configurations, it is also solvable for a certain range of constraint density $α$ and threshold $κ$. We provide in this paper a statistical measure probing sequences of solutions, where two consecutive elements shares a strong overlap. After simplifications, we test its predictions by comparing it to Monte-Carlo simulations. We obtain good agreement and show that connected states with a Markovian correlation profile can fully decorrelate from their initialization only for $κ>κ_{\rm no-mem.\, state}$ ($κ_{\rm no-mem.\, state}\sim \sqrt{0.91\log(N)}$ for $α=0.5$ and $N$ being the dimension of the problem). For $κ<κ_{\rm no-mem.\, state}$, we show that decorrelated sequences still exist but have a non-trivial correlations profile. To study this regime we introduce an $Ansatz$ for the correlations that we label as the nested Markov chain.
title How to escape atypical regions in the symmetric binary perceptron: a journey through connected-solutions states
topic Disordered Systems and Neural Networks
Statistical Mechanics
Mathematical Physics
url https://arxiv.org/abs/2408.04479