Whether Mirror and Conceptual Neurons are Myths? Sparse vs. Distributed Neuronal Representations

Abstract

ABSTRACT Multi-layer neural networks, mirror neurons, and gnostic neurons are concepts that assign neural representations to mental representations of percepts and inner sensations. However, none of these approaches alone can explain the higher mental functions, which we observe in natural minds from the third and first-person perspectives through introspection. Recent concepts of preservation of chemical traces of sensory stimuli and hierarchical structures of postsynaptic associations represented by specifically organized groups of neurons combine these concepts and effectively explain much more complex mental functions. To find an operative model and understand how knowledge in the mind creates conscious sensations, we explain how perceptions, sensory impressions, and environment models gain their neural representations. It was pointed out ways to detect the similarity of structures representing previously remembered patterns to the mental and neuronal representations of new perceptions, ways of their associations, and principles of information processing. Supplemented, presented in earlier works, concepts of competition of representations stimulation and factors stimulating their action explain the mind’s complex functions, including speech production and recognition. We postulate that using new methods of modelling the neural network’s functions through the parallel physical process allows creating a physical model of natural and artificial, conscious, intelligent minds.