Abstract "... We speculate on the significance of polychrony to the theory of neuronal group selection (TNGS, neural Darwinism), cognitive neural computations, binding and gamma rhythm, mechanisms of attention, and consciousness as "attention to memories. ..."


"We hypothesize that polychronous groups could represent memories and experience."


"Sleep states: The network can switch randomly between different states. Some of them correspond to "vigilance" with gamma oscillations, and others resemble "sleep" states, similar to the one in Figure 5."


"the network was prone to epileptic seizures, which eventually lead to uncontrolled, completely synchronized activity."



"The group is said to activate when more than 50% of its neurons polychronize, that is, fire with the prescribed spike-timing pattern with �}1 ms jitter, as in Figure 6."


We might say that the stimulus is the focus of
attention. If two or more stimuli are present, then activation of groups representing
one stimulus essentially precludes the other stimuli from being
attended.


One can say that the network �gthinks�h of the stimulus.that is,
it pays attention to the memory of the stimulus. Such �gthinking�h resembles
�gexperiencing�h the stimulus."


--------------------------------------------------------------------------------------------------

To explore the issue of spike timing in networks with conduction delays,we simulated an anatomically realisticmodel consisting of 100,000 cortical spiking neurons having receptors with AMPA, NMDA, GABAA, andGABAB kinetics and long-term and short-term synaptic plasticity (Izhikevich, Gally, & Edelman, 2004).

Izhikevich, E. M., Gally, J. A., & Edelman, G. M. (2004). 
Spike-timing dynamics of neuronal groups. 
Cerebral Cortex, 14, 933.944. => done


We do not model modifiable delays (Huning, Glunder, & Palm, 1998; Eurich, Pawelzik, Ernst, Cowan, & Milton, 1999)

Eurich, C., Pawelzik, K., Ernst, U., Cowan, J., & Milton, J. (1999). 
Dynamics of selforganazed delay adaptation.
 Phys. Rev. Lett., 82, 1594.1597. => done

Huning,H., Glunder, H.,&Palm, G. (1998). 
Synaptic delay learning in pulse-coupled neurons.
 Neural Computation, 10, 555.565. => not-available



If b, c, and d are sensory neurons driven by an external input, then the simple circuit in Figure 2 can recognize and classify simple spatiotemporal patterns (Hopfield, 1995; Seth, Mckinstry, Edelman, & Krichmar, 2004b.)

Seth,A. K., McKinstry, J. L., Edelman, G. M.,&Krichmar, J. L. (2004b). 
Visual binding through reentrant connectivity and dynamic synchronization in a brain-based device.
 Cerebral Cortex, 14, 1185.1199. => done

--------------------------------------------------------------------------------------------------
In the simulation above, no coherent external input to the system was present. As a result, random groups emerge; that is, the network generates random memories not related to any previous experience.
--------------------------------------------------------------------------------------------------






--------------------------------------------------------------------------------------------------
A careful measurement of axonal conduction delays in the
mammalian neocortex (Swadlow 1985, 1988, 1992) showed that they could
be as small as 0.1 ms and as large as 44 ms, depending on the type and
location of the neurons. ... Why would
the brain maintain different delays with such precision if spike timing were
not important?

In this letter,weargue that an infinite dimensionality of spiking networks
with axonal delays is not a nuisance but an immense advantage that results
in an unprecedented information capacity. In particular, there are stable
firing patterns that are not possible without the delays.

In this letterwe present aminimal model that captures the essence of this
phenomenon.

Synaptic connections among neurons have fixed conduction
delays, which are random integers between 1 ms and 20 ms.