Mean-Field Theory of Irregularly Spiking Neuronal Populations and Working Memory in Recurrent Cortical Networks - Computational Neuroscience: A Comprehensive Approach - page 470

40

EXC
INH

30

20

10

0

1.2

EXC
INH

1.1

1

30

20

10 m

s

V

0

500 1000 1500 2000

t (ms)

Figure 15.8
Bistability in a balanced multi-columnar cortical circuit. (A) . Temporal evolution of
the firing rate from the excitatory and inhibitory sub-populations of a column. At
t 500 ms, a transient excitatory input was applied to both sub-populations. The

elevated activity state in response to this input outlasts the stimulus offset. Note
the elevated firing rate of the inhibitory sub-population also. (B) . Same as above
for the CV of the two sub-populations. The CV increases with the firing rate. (C) .
The figure shows the quantities m V =

=

g L and s V = √ t m s C / C m of the neurons in

the excitatory sub-population. They correspond to the mean and standard deviation
of the current, but are expressed in mV to facilitate comparison with the distance
between V th and V L , equal to 20 mV (dashed line). The mean input current remains
essentially the same for both the resting state and persistent state, regardless of their
very different firing rates. The increase in firing rate in response to the stimulus is due
to an increase in the amplitude of the fluctuating component of the current, hence the
increase in CV above. In this network, both stable states are in the balanced regime.

m C /