Neural network comparing two-rate-encoded inputs entering in parallel.

Presented computational model of a neural network is able to compare two regular input frequencies. The comparison is based on detection of inter-spike interval differences of the two frequencies. This detection is continuous and the network dynamically changes its output according to the changes in the input frequencies. The entire network is composed of biologically plausible parts. A combination of such simple comparators might be involved in the information processing in the central nervous system.