Abstract:This paper investigates the controllability properties of a general class of recurrent neural networks that are widely used for hypothesis generation in theoretical neuroscience, including the modeling of large-scale human brain dynamics. Our study focuses on the control synthesis of such networks using constant and piecewise constant inputs, motivated by emerging applications in non-invasive neurostimulation such as transcranial direct current stimulation (tDCS). The neural network model considered is a continuous Hopfield-type system with nonlinear activation functions and arbitrary input matrices representing interactions among multiple brain regions. Our main contribution is the formulation and solution of a control synthesis problem for these nonlinear systems. We provide a proper generalization of the variation of the constants formula that constitutes a novel representation of the system's state trajectory. This representation admits a verifiable condition on the existence of the constant control input to solve a short-time two-point boundary value problem in the state space. This formulation admits a synthesis for the input in question, which can be realized using modern algorithmic optimization tools. In the case of linear activation functions, this analysis and synthesis reduces to the verification of algebraic conditions on the system matrices. Simulation results are presented to illustrate the theoretical findings and demonstrate the efficacy of the proposed control strategies. These results offer a novel control synthesis for an important class of neural network models that may, in turn, enable the design of brain stimulation protocols to modulate whole-brain activity in therapeutic and cognitive enhancement applications.

Effect of Constant External Input on Continuous Attractors Shifting

Continuous Attractors of Discrete-Time Recurrent Neural Networks

Dynamical Properties of Continuous Attractor Neural Network with Background Tuning

Dynamic shift mechanism of continuous attractors in a class of recurrent neural networks

Continuous Attractors Of A Class Of Recurrent Neural Networks Without Lateral Inhibition

Coexistence of Multiple Continuous Attractors for Lower-Ordered Neural Networks

Analysis of Continuous Attractors for 2-D Linear Threshold Neural Networks

Continuous Attractors and Population Decoding Multiple-Peaked Activity

The Tracking Speed of Continuous Attractors

Representations of Continuous Attractors of Recurrent Neural Networks

Dynamics and computation of continuous attractors.

Continuous Attractors of Lotka-Volterra Recurrent Neural Networks with Infinite Neurons

Computing with Continuous Attractors: Stability and Online Aspects

On the control of recurrent neural networks using constant inputs

Nontrivial Global Attractors in 2-D Multistable Attractor Neural Networks

Representations Of Continuous Attractors Of Discrete-Time Cellular Neural Networks

The Effect of Coefficients on the Continuous Attractors in Coupled Highway Neural Networks

Non-equilibrium statistical mechanics of continuous attractors

Dynamics Of Neural Networks With Continuous Attractors

Coexistence of continuous attractors with different dimensions for neural networks

Continuous Attractors of a Class of Neural Networks with a Large Number of Neurons