Phase-locked states in a spiking neural network model with a context-dependent connectivity

In this paper, we propose a spiking neural network model with Hebbian connectivity that can adjust its activity to input stimuli. The implemented architecture is suitable for the task of recognizing binary images encoded using in-phase and anti-phase oscillations relative to the global clock signal. The phase locking effect makes it possible to achieve cluster synchronization of neurons (both at the input and at the output layer). In this work, a context-dependent algorithm for initiating oscillations of spiking interneurons, which are in an excitable mode, is implemented. As a result, only a fraction of the Hebb connections are employed for the pattern retrieval from memory. The biologically relevant Hodgkin-Huxley-Mainen model is utilized as neurons.