Relationship between spike irregularity and neural network dynamics

It is well known that interspike intervals (ISI) in the cortex have high values of the coefficient of variation (Cv). The purpose of this study is to clarify the relation between spike irregularity and network dynamics. We observe the irregularity of interspike intervals when we change the values of two parameters τ and τ_ref which strongly affect the network dynamics in a spiking neural network model. The results show that spike irregularity depends on the neural network dynamics as follows: When τ (τ_ref) is larger (smaller), the irregularity increases and the stability of firing rates decreases. On the other hand, these changes are unrelated to the largest Lyapunov exponent, an indicator of instability of an orbit in the state space.