Cell Assemblies

A concept (like Dog) is represented in your brain by a reverberating circuit of neurons called a Cell Assembly.
That's a pretty well supported theory from Don Hebb in 1949.
A short term memory is activated when that circuit starts to fire at an elevated rate. The CA ignites.
If you have the circuit, it's a Long Term Memory.
I'm a CA guy.
I've got a neural model of CAs called a binary Cell Assembly.
Roughly, it's a well connected set of neurons that can persist indefinitely (or for a long time with adaptation) once ignited.
It's not a great model of biological CAs, but I'm not aware of a great model.
For that matter, I'm not aware of good biological data that describes a good CA. (It's hard to see what a lot of real biological neurons are doing in real time.)
What's cool about these binary CAs is that you can use them to implement a lot of programs (e.g. finite state automata) with not many neurons.
So, you can write code that runs simulated (or emulated neuromorphic) neurons to do anything.
These work well for distributed processing.
They are not necessarily the correct model of the brain or even a little part of it.