Stable fixed points of combinatorial threshold-linear networks

DOI10.1016/J.AAM.2023.102652MaRDI QIDQ6138049zbMATH OpenOpenAlexFDO

Authors Carina Curto, Jesse Geneson, Katherine Morrison

Publication date 16 January 2024

Published in Advances in Applied Mathematics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1909.02947

zbMATH Keywords

cliquesCollatz-Wielandt formulaattractor neural networksstable fixed pointsthreshold-linear networks

Mathematics Subject Classification ID

Boundary value problems on graphs and networks for ordinary differential equations (34B45) Fixed points and periodic points of dynamical systems; fixed-point index theory; local dynamics (37C25) Stability theory for smooth dynamical systems (37C75) Dynamical systems involving maps of trees and graphs (37E25)

Abstract: Combinatorial threshold-linear networks (CTLNs) are a special class of recurrent neural networks whose dynamics are tightly controlled by an underlying directed graph. Recurrent networks have long been used as models for associative memory and pattern completion, with stable fixed points playing the role of stored memory patterns in the network. In prior work, we showed that target-free cliques of the graph correspond to stable fixed points of the dynamics, and we conjectured that these are the only stable fixed points possible. In this paper, we prove that the conjecture holds in a variety of special cases, including for networks with very strong inhibition and graphs of size $nleq4$. We also provide further evidence for the conjecture by showing that sparse graphs and graphs that are nearly cliques can never support stable fixed points. Finally, we translate some results from extremal combinatorics to upper bound the number of stable fixed points of CTLNs in cases where the conjecture holds.

Recommendations

• Fixed Points of Competitive Threshold-Linear Networks
• Pattern completion in symmetric threshold-linear networks
• Sequential Attractors in Combinatorial Threshold-Linear Networks
• Permitted and Forbidden Sets in Symmetric Threshold-Linear Networks
• Attractive periodic sets in discrete-time recurrent networks (with emphasis on fixed-point stability and bifurcations in two-neuron networks)

Cites work

• scientific article; zbMATH DE number 1164156 (Why is no real title available?)
• Encoding binary neural codes in networks of threshold-linear neurons
• Fixed Points of Competitive Threshold-Linear Networks
• Flexible memory networks
• Graph Rules for Recurrent Neural Network Dynamics
• Modeling Brain Function
• Neural networks and physical systems with emergent collective computational abilities
• On cliques in graphs
• On cliques in graphs
• On cliques in graphs
• Optimal control problems for a class of non-linear evolution equations
• Pattern completion in symmetric threshold-linear networks
• Permitted and Forbidden Sets in Symmetric Threshold-Linear Networks
• Rat Prefrontal Cortex Inactivations during Decision Making Are Explained by Bistable Attractor Dynamics
• Selectively Grouping Neurons in Recurrent Networks of Lateral Inhibition
• Sequential Attractors in Combinatorial Threshold-Linear Networks
• The maximum number of cliques in dense graphs
• The worst-case time complexity for generating all maximal cliques and computational experiments

This page was built for publication: Stable fixed points of combinatorial threshold-linear networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6138049)