activity
20182020
collaborators

5 papers

cond-mat.dis-nn2020

On the number of limit cycles in diluted neural networks

Sungmin Hwang, Enrico Lanza, Giorgio Parisi +3

We consider the storage properties of temporal patterns, i.e. cycles of finite lengths, in neural networks represented by (generally asymmetric) spin glasses defined on random grap…

cond-mat.dis-nn2019

Force balance controls the relaxation time of the gradient descent algorithm in the satisfiable phase

Sungmin Hwang, Harukuni Ikeda

We numerically study the relaxation dynamics of the single layer perceptron with the spherical constraint. This is the simplest model of neural networks and serves a prototypical m…

hep-th2018

Poincaré invariance in NRQCD and pNRQCD revisited

Matthias Berwein, Nora Brambilla, Sungmin Hwang +1

We investigate how fields transform under the Poincaré group in nonrelativistic effective field theories of QCD. In constructing these transformations, we rely only on symmetries a…

cond-mat.dis-nn2018

On the number of limit cycles in asymmetric neural networks

Sungmin Hwang, Viola Folli, Enrico Lanza +3

The comprehension of the mechanisms at the basis of the functioning of complexly interconnected networks represents one of the main goals of neuroscience. In this work, we investig…

cond-mat.dis-nn2018

Jamming in multilayer supervised learning models

Silvio Franz, Sungmin Hwang, Pierfrancesco Urbani

Critical jamming transitions are characterized by an astonishing degree of universality. Analytic and numerical evidence points to the existence of a large universality class that…