Topological and Dynamical Complexity of Random Neural Networks

Abstract

Random neural networks are dynamical descriptions of randomly interconnected neural units. These show a phase transition to chaos as a disorder parameter is increased. The microscopic mechanisms underlying this phase transition are unknown, and similarly to spin-glasses, shall be fundamentally related to the behavior of the system. In this Letter we investigate the explosion of complexity arising near that phase transition. We show that the mean number of equilibria undergoes a sharp transition from one equilibrium to a very large number scaling exponentially with the dimension on the system. Near criticality, we compute the exponential rate of divergence, called topological complexity. Strikingly, we show that it behaves exactly as the maximal Lyapunov exponent, a classical measure of dynamical complexity. This relationship unravels a microscopic mechanism leading to chaos which we further demonstrate on a simpler class of disordered systems, suggesting a deep and underexplored link between topological and dynamical complexity.

References in corpus (2)

• Heterogeneous connections induce oscillations in large scale networks
• Eigenvalue spectra of asymmetric random matrices for multi-component neural networks

Cited by in corpus (62)

• Extreme value statistics of correlated random variables: a pedagogical review
• Transition to chaos in random neuronal networks
• Distributions of covariances as a window into the operational regime of neuronal networks
• Two types of criticality in the brain
• Nonlinear Analogue of the May-Wigner Instability Transition
• Dynamics of random recurrent networks with correlated low-rank structure
• Optimal sequence memory in driven random networks
• Intrinsically-generated fluctuating activity in excitatory-inhibitory networks
• Correlations between synapses in pairs of neurons slow down dynamics in randomly connected neural networks
• Hopfield Networks is All You Need
• Complex energy landscapes in spiked-tensor and simple glassy models: ruggedness, arrangements of local minima and phase transitions
• Determination of the edge of criticality in echo state networks through Fisher information maximization
• Turning intractable counting into sampling: computing the configurational entropy of three-dimensional jammed packings
• Generalized Lotka-Volterra equations with random, non-reciprocal interactions: the typical number of equilibria
• Mean field limit for disordered diffusions with singular interactions
• Neural Field Models: A mathematical overview and unifying framework
• Does the brain behave like a (complex) network? I. Dynamics
• Exponential number of equilibria and depinning threshold for a directed polymer in a random potential
• Large deviations, dynamics and phase transitions in large stochastic heterogeneous neural networks
• Effect of dilution in asymmetric recurrent neural networks
• Hessian spectrum at the global minimum of high-dimensional random landscapes
• Topology trivialization transition in random non-gradient autonomous ODE's on a sphere
• Transient chaotic dimensionality expansion by recurrent networks
• Conditioned stochastic particle systems and integrable quantum spin systems
• Generic Properties of Random Gene Regulatory Networks
• Transition from Gaussian to non-Gaussian fluctuations for mean-field diffusions in spatial interaction
• Nonlinearity-generated Resilience in Large Complex Systems
• Counting equilibria in a random non-gradient dynamics with heterogeneous relaxation rates
• Quenched complexity of equilibria for asymmetric Generalized Lotka-Volterra equations
• Extended Anderson Criticality in Heavy-Tailed Neural Networks
• Stability of large complex systems with heterogeneous relaxation dynamics
• Distribution of rare saddles in the $p$-spin energy landscape
• Consequences of Dale's law on the stability-complexity relationship of random neural networks
• Activated dynamics: an intermediate model between REM and p-spin
• Manifolds in high dimensional random landscape: complexity of stationary points and depinning
• Dynamically selected steady states and criticality in non-reciprocal networks
• Spectrum of non-Hermitian deep-Hebbian neural networks
• Kac-Rice fixed point analysis for single- and multi-layered complex systems
• Instability to a heterogeneous oscillatory state in randomly connected recurrent networks with delayed interactions
• Energy Landscape of the Finite-Size Mean-field 2-Spin Spherical Model and Topology Trivialization
• Saturation of number variance in embedded random matrix ensembles
• Balanced condition in networks leads to Weibull statistics
• An optimization-based equilibrium measure describes non-equilibrium steady state dynamics: application to edge of chaos
• Manifolds pinned by a high-dimensional random landscape: Hessian at the global energy minimum
• Counting equilibria of large complex systems by instability index
• When is the average number of saddle points typical?
• Statistics of Stationary Points of Random Finite Polynomial Potentials
• Exact finite-size scaling for the random-matrix representation of bond percolation on square lattice
• Optimal system size for complex dynamics in random neural networks near criticality
• Irreversibility in Non-reciprocal Chaotic Systems
• Right large deviation principle for the top eigenvalue of the sum or product of invariant random matrices
• Evaluating the statistical similarity of neural network activity and connectivity via eigenvector angles
• Discontinuous transition to chaos in a canonical random neural network
• Interaction-correlated random matrices
• Regular graphs maximize the variability of random neural networks
• Robust quantum metrology with random Majorana constellations
• Freezing chaos without synaptic plasticity
• Synaptic plasticity alters the nature of chaos transition in neural networks
• Plastic systemic inhibition controls amplitude while allowing phase pattern in a stochastic neural field model
• Critical behavior of a phase transition in the dynamics of interacting populations
• Phase transitions from linear to nonlinear information processing in neural networks
• Asymmetry in repeated isotropic rotations