Functional differentiations in evolutionary reservoir computing networks

We propose an extended reservoir computer that shows the functional differentiation of neurons. The is developed to enable changing internal using evolutionary dynamics, and we call it computer. To develop neuronal units show specificity, depending on input information, dynamics should be controlled produce contracting after expanding dynamics. Expanding magnifies difference while contributes forming clusters thereby producing multiple attractors. simultaneous appearance both indicates existence chaos. In contrast, sequential these during finite time intervals may induce differentiations. this paper, how specific are yielded in