Tsodyks: Duality of cortical neurons as ``binary'' and ``analog'' computational elements

Weizmann Institute

We analyze some computational implications of short- and long- term modifications in synaptic transmission between neocortical pyramidal neurons. We argue that a cortical neuron receives from its presynaptic sources a very rich information about their activity state, including ``binary'' signals about their stationary activity level, and ``analog'' signals about the timings and amplitudes of its variations.