Abstract Details

Cortex-like Complex Systems: What Occurs Within?  Peter Grindrod , Christopher Lester, Mathematical Institute University Of Oxford (Marhemticak Instriture, University of Oxford, Oxford, OXON United Kingdom)   C19

We consider a class of cortex-like complex systems in the form of strongly connected, directed, networks-of-networks with spiking (excitatory and refractive) dynamics at each of the nodes, together with time-lags, that are non trivial compared to the timescale of a single spike, associated with each of the directed edges. These systems can process various incoming stimulations, in the form of forcing at specific nodes, by producing whole-system dynamical responses. The outer network is sparsely connected, while the many inner networks, called {it modules}, are relatively densely connected. We discuss a generic class of such systems with up to 10B nodes (neurones) simulating the human cerebral cortex. Such a dynamical system's responses to a rather wide range of distinct stimulations may be classified into a number of latent, internal, dynamical patterns (characterised both in time and across the network). Such patterns are internal and subjective to the system, and their appearance would constrain its immediate responses to further immediate simulating inputs. The patterns may focus and bias (inhibit or encourage) short-term responses resulting in an evolutionary advantage, called "fast thinking" , by decision-making within a reduced domain of possible perceptions (a biased set of possibilities). We illustrate how such internal (latent) patterns may be shown to exist within very large-scale simulations for a rather wide and appropriate class of complex systems. Hence they must also exist within the human brain. The existence of, and internal experiences furnished by, these latent patterns are part of "what it is like" to be a human brain. There is nothing more.