Global brain circuitry control of behavior emerging from self-governing vector field dynamics in subnetworks

Abstract

Behavior ultimately depends on the spatiotemporal patterns of the neuron population activity across the brain. Here we address the issue of how the evolving patterns of population activity can be governed by the intrinsically available mechanisms within the brain. We show how the control of the evolving activity can be represented by a high-dimensional vector field, which is an emergent effect of the integrative effects afforded by the membrane capacitance of the neurons and the weights of the synaptic connections between them. For each subnetwork of the brain, its intrinsic connectivity defines the structure of a lower-dimensional vector field with a local attractor point, towards which the subnetwork activity is constantly drawn. We show that other subnetworks, defined by having a degree of independence from but an impact on the first subnetwork, will constantly move the location of the local attractor point, causing the activity in the first subnetwork to constantly chase its own tail. We show how this principle can explain how minor differences in the corticospinal control signal can produce a variety of movement patterns through the spinal interneuron circuitry. At the global level, we show how the cortical neuron population can be thought of as many concatenated subnetworks that produce diverse dynamic evolutions across the cortical neuron population globally. This operational principle can produce a dynamic population activity reminiscent of that observed across the brain in vivo and explain the foundational mechanisms underlying autonomous cortical control of its own activity evolution.