Extended Anderson Criticality in Heavy-Tailed Neural Networks

被引:6
|
作者
Wardak, Asem [1 ]
Gong, Pulin [1 ]
机构
[1] Univ Sydney, Sch Phys, Sydney, NSW 2006, Australia
基金
澳大利亚研究理事会;
关键词
RANDOM MATRICES; EIGENVECTORS; BRAIN; CHAOS;
D O I
10.1103/PhysRevLett.129.048103
中图分类号
O4 [物理学];
学科分类号
0702 ;
摘要
We investigate the emergence of complex dynamics in networks with heavy-tailed connectivity by developing a non-Hermitian random matrix theory. We uncover the existence of an extended critical regime of spatially multifractal fluctuations between the quiescent and active phases. This multifractal critical phase combines features of localization and delocalization and differs from the edge of chaos in classical networks by the appearance of universal hallmarks of Anderson criticality over an extended region in phase space. We show that the rich nonlinear response properties of the extended critical regime can account for a variety of neural dynamics such as the diversity of timescales, providing a computational advantage for persistent classification in a reservoir setting.
引用
收藏
页数:6
相关论文
共 50 条
  • [1] Heavy-tailed configuration models at criticality
    Dhara, Souvik
    van der Hofstad, Remco
    van Leeuwaarden, Johan S. H.
    Sen, Sanchayan
    [J]. ANNALES DE L INSTITUT HENRI POINCARE-PROBABILITES ET STATISTIQUES, 2020, 56 (03): : 1515 - 1558
  • [2] The parabolic Anderson model with heavy-tailed potential
    Moerters, Peter
    [J]. SURVEYS IN STOCHASTIC PROCESSES, 2011, : 67 - 85
  • [3] Heavy-Tailed Regularization of Weight Matrices in Deep Neural Networks
    Xiao, Xuanzhe
    Li, Zeng
    Xie, Chuanlong
    Zhou, Fengwei
    [J]. ARTIFICIAL NEURAL NETWORKS AND MACHINE LEARNING, ICANN 2023, PART X, 2023, 14263 : 236 - 247
  • [4] Measure for characterizing heavy-tailed networks
    Hill, Sam A.
    [J]. PHYSICAL REVIEW RESEARCH, 2021, 3 (02):
  • [5] Estimation of extreme quantiles from heavy-tailed distributions with neural networks
    Allouche, Michael
    Girard, Stephane
    Gobet, Emmanuel
    [J]. STATISTICS AND COMPUTING, 2024, 34 (01)
  • [6] Estimation of extreme quantiles from heavy-tailed distributions with neural networks
    Michaël Allouche
    Stéphane Girard
    Emmanuel Gobet
    [J]. Statistics and Computing, 2024, 34
  • [7] Edge of Chaos and Avalanches in Neural Networks with Heavy-Tailed Synaptic Weight Distribution
    Kusmierz, Lukasz
    Ogawa, Shun
    Toyoizumi, Taro
    [J]. PHYSICAL REVIEW LETTERS, 2020, 125 (02)
  • [8] Packet reordering in networks with heavy-tailed delays
    Marc Lelarge
    [J]. Mathematical Methods of Operations Research, 2008, 67 : 341 - 371
  • [9] Robust sequential learning of feedforward neural networks in the presence of heavy-tailed noise
    Vukovic, Najdan
    Miljkovic, Zoran
    [J]. NEURAL NETWORKS, 2015, 63 : 31 - 47
  • [10] Packet reordering in networks with heavy-tailed delays
    Lelarge, Marc
    [J]. MATHEMATICAL METHODS OF OPERATIONS RESEARCH, 2008, 67 (02) : 341 - 371