Onset of chaotic dynamics in neural networks

Gianbiagio Curato*, Antonio Politi

*Corresponding author for this work

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

A neural-network model is proposed as a test bed for the characterization of the chaotic dynamics emerging in a context where the coupling is, on the average, neither excitatory nor inhibitory. The proposed discrete-time model generalizes within a single framework two different setups previously studied in the literature. With the help of theoretical mean field arguments and numerical simulations on GPUs, we characterize the transition and show that the chaotic dynamics is extensive (i.e., that the number of active degrees of freedom is proportional to the network size) from the very beginning. Besides the coupling strength, two parameters play a crucial role: (1) one controls the local dissipation and determines the shape of the initial part of the Lyapunov spectrum as well as the shape of the correlation function; (2) the other, which corresponds to the amplitude of an effective random field, determines the nature of the transition.

Original languageEnglish
Article number042908
Pages (from-to)1-7
Number of pages7
JournalPhysical Review. E, Statistical, Nonlinear and Soft Matter Physics
Volume88
Issue number4
DOIs
Publication statusPublished - 14 Oct 2013

Fingerprint

Neural Networks (Computer)
Chaotic Dynamics
Neural Networks
Lyapunov Spectrum
Discrete-time Model
test stands
Neural Network Model
Testbed
Mean Field
Random Field
Correlation Function
Two Parameters
Dissipation
emerging
dissipation
degrees of freedom
Degree of freedom
Directly proportional
Numerical Simulation
Generalise

ASJC Scopus subject areas

  • Condensed Matter Physics
  • Statistical and Nonlinear Physics
  • Statistics and Probability
  • Medicine(all)

Cite this

Onset of chaotic dynamics in neural networks. / Curato, Gianbiagio; Politi, Antonio.

In: Physical Review. E, Statistical, Nonlinear and Soft Matter Physics, Vol. 88, No. 4, 042908, 14.10.2013, p. 1-7.

Research output: Contribution to journalArticle

@article{f937aa5b57be4d3cb25aa1a52551a88e,
title = "Onset of chaotic dynamics in neural networks",
abstract = "A neural-network model is proposed as a test bed for the characterization of the chaotic dynamics emerging in a context where the coupling is, on the average, neither excitatory nor inhibitory. The proposed discrete-time model generalizes within a single framework two different setups previously studied in the literature. With the help of theoretical mean field arguments and numerical simulations on GPUs, we characterize the transition and show that the chaotic dynamics is extensive (i.e., that the number of active degrees of freedom is proportional to the network size) from the very beginning. Besides the coupling strength, two parameters play a crucial role: (1) one controls the local dissipation and determines the shape of the initial part of the Lyapunov spectrum as well as the shape of the correlation function; (2) the other, which corresponds to the amplitude of an effective random field, determines the nature of the transition.",
author = "Gianbiagio Curato and Antonio Politi",
note = "A.P. thanks Alessandro Torcini for helpful discussions about the nature of the balanced state. G.C. thanks Alessandro Profeti for helpful discussions about the innovative features of the complex microprocessor GK104(Kepler).",
year = "2013",
month = "10",
day = "14",
doi = "10.1103/PhysRevE.88.042908",
language = "English",
volume = "88",
pages = "1--7",
journal = "Physical Review. E, Statistical, Nonlinear and Soft Matter Physics",
issn = "1539-3755",
publisher = "AMER PHYSICAL SOC",
number = "4",

}

TY - JOUR

T1 - Onset of chaotic dynamics in neural networks

AU - Curato, Gianbiagio

AU - Politi, Antonio

N1 - A.P. thanks Alessandro Torcini for helpful discussions about the nature of the balanced state. G.C. thanks Alessandro Profeti for helpful discussions about the innovative features of the complex microprocessor GK104(Kepler).

PY - 2013/10/14

Y1 - 2013/10/14

N2 - A neural-network model is proposed as a test bed for the characterization of the chaotic dynamics emerging in a context where the coupling is, on the average, neither excitatory nor inhibitory. The proposed discrete-time model generalizes within a single framework two different setups previously studied in the literature. With the help of theoretical mean field arguments and numerical simulations on GPUs, we characterize the transition and show that the chaotic dynamics is extensive (i.e., that the number of active degrees of freedom is proportional to the network size) from the very beginning. Besides the coupling strength, two parameters play a crucial role: (1) one controls the local dissipation and determines the shape of the initial part of the Lyapunov spectrum as well as the shape of the correlation function; (2) the other, which corresponds to the amplitude of an effective random field, determines the nature of the transition.

AB - A neural-network model is proposed as a test bed for the characterization of the chaotic dynamics emerging in a context where the coupling is, on the average, neither excitatory nor inhibitory. The proposed discrete-time model generalizes within a single framework two different setups previously studied in the literature. With the help of theoretical mean field arguments and numerical simulations on GPUs, we characterize the transition and show that the chaotic dynamics is extensive (i.e., that the number of active degrees of freedom is proportional to the network size) from the very beginning. Besides the coupling strength, two parameters play a crucial role: (1) one controls the local dissipation and determines the shape of the initial part of the Lyapunov spectrum as well as the shape of the correlation function; (2) the other, which corresponds to the amplitude of an effective random field, determines the nature of the transition.

UR - http://www.scopus.com/inward/record.url?scp=84886053023&partnerID=8YFLogxK

U2 - 10.1103/PhysRevE.88.042908

DO - 10.1103/PhysRevE.88.042908

M3 - Article

VL - 88

SP - 1

EP - 7

JO - Physical Review. E, Statistical, Nonlinear and Soft Matter Physics

JF - Physical Review. E, Statistical, Nonlinear and Soft Matter Physics

SN - 1539-3755

IS - 4

M1 - 042908

ER -