Excitable networks for finite state computation with continuous time recurrent neural networks

Continuous time recurrent neural networks (CTRNN) are systems of coupled ordinary differential equations that are simple enough to be insightful for describing learning and computation, from both biological and machine learning viewpoints. We describe a direct constructive method of realising finite state input-dependent computations on an arbitrary directed graph. The constructed system has an excitable network attractor whose dynamics we illustrate with a number of examples. The resulting CTRNN has intermittent dynamics: trajectories spend long periods of time close to steady-state, with rapid transitions between states. Depending on parameters, transitions between states can either be excitable (inputs or noise needs to exceed a threshold to induce the transition), or spontaneous (transitions occur without input or noise). In the excitable case, we show the threshold for excitability can be made arbitrarily sensitive.

• Publication year

  2020

• Venue

  Biological cybernetics

• Publication date

  2020-12-08

• Fields of study

  Medicine, Computer Science, Mathematics

• Identifiers
  DOI 10.1007/s00422-021-00895-5arXiv 2012.04129PMID 34608540PMCID 8589808
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar, PubMed

• No claims are published for this paper.

• No concepts are published for this paper.

• Sequential dynamics of complex networks in mind: Consciousness and creativity

  2020cited by this paper
• The dynamics of continuous-time recurrent neural networks and their relevance to episodic memory

  2019cited by this paper
• Before and beyond the Wilson-Cowan equations.

  2019cited by this paper
• Sensitive Finite-State Computations Using a Distributed Network With a Noisy Network Attractor

  2018cited by this paper
• Interpreting Recurrent Neural Networks Behaviour via Excitable Network Attractors

  2018cited by this paper
• Almost Complete and Equable Heteroclinic Networks

  2018cited by this paper
• Sequences by Metastable Attractors: Interweaving Dynamical Systems and Experimental Data

  2017cited by this paper
• A reservoir computing model of episodic memory

  2016cited by this paper
• Bifurcations in the Wilson-Cowan Equations with Nonsmooth Firing Rate

  2015cited by this paper
• Nonlinear Dynamics and Chaos: With Applications to Physics, Biology, Chemistry, and Engineering

  2015cited by this paper
• Designing Heteroclinic and Excitable Networks in Phase Space Using Two Populations of Coupled Cells

  2015cited by this paper
• Theory of Input Driven Dynamical Systems

  2012cited by this paper
• Generation and reshaping of sequences in neural systems

  2006cited by this paper
• On the origin of reproducible sequential activity in neural circuits.

  2004cited by this paper
• Exploring the T-Maze: Evolving Learning-Like Robot Behaviors Using CTRNNs

  2003cited by this paper
• Heteroclinic Contours in Neural Ensembles and the Winnerless Competition Principle

  2003cited by this paper
• An Evolutionary Ecological Approach to the Study of Learning Behavior Using a Robot-Based Model

  2002cited by this paper
• A class of piecewise linear differential equations arising in biological models

  2002cited by this paper
• Dynamical encoding by networks of competing neuron groups: winnerless competition.

  2001cited by this paper
• On the Dynamics of Small Continuous-Time Recurrent Neural Networks

  1995influential reference
• A competition between heteroclinic cycles

  1994cited by this paper
• Sequential Behavior and Learning in Evolved Dynamical Neural Networks

  1994cited by this paper
• Original Contribution: Approximation of dynamical systems by continuous time recurrent neural networks

  1993cited by this paper
• Structurally stable heteroclinic cycles

  1988cited by this paper
• “Neural” computation of decisions in optimization problems

  1985influential reference
• Convection in a Rotating Layer: A Simple Case of Turbulence

  1980cited by this paper
• Nonlinear Aspects of Competition Between Three Species

  1975cited by this paper
• Excitatory and inhibitory interactions in localized populations of model neurons.

  1972cited by this paper

• Between Behaviors: Comparison of Two Dynamical Models of Behavioral Switching for C. Elegans Locomotion

  2026influential citation
• An analog-electronic implementation of a harmonic oscillator recurrent neural network

  2025cites this paper
• Arbitrary sensitive transitions in recurrent neural networks

  2024cites this paper
• A Continuous Time Dynamical Turing Machine

  2024influential citation
• Network attractors and nonlinear dynamics of neural computation.

  2023cites this paper
• On the Dynamics of Learning Time-Aware Behavior with Recurrent Neural Networks

  2023cites this paper
• Stability of heteroclinic cycles in rings of coupled oscillators

  2022cites this paper
• English Lexical Analysis System of Machine Translation Based on Simple Recurrent Neural Network

  2022cites this paper
• Stability of heteroclinic cycles in ring graphs.

  2022cites this paper