Improved Cell Assemblies

I think the big thing to work on is improved CAs.
If we get the grant, most of my first year will be devoted to improved CAs.
Even if we don't get this grant (or another one), I'm going to spend a lot of time on this.
A main point of CAs is that they're the neural basis of pyschological concepts.
Unfortunately, there doesn't seem to be much work on modelling basic psychological phenomena in CAs that emerge from neurons.
The problems here include variable activation, semantic nets, fractionation and recruitment, and cognitive mapping.
Variable activation includes gradual decay of short term memorys over time (which we kind of get); reactivation and timing; associative memory; and variable input.
Possible solutions include multiple neuron types, topological restrictions and inhibitory feedback mechanisms.