Short Term Stable States
- This memory CA is a stable state.
- When a CA ignites, the neurons in it are more likely to be
on than neurons outside of it.
- Hopefully, the difference is extreme with say 30% of the neurons
in the CA firing, and only 1% of the neurons outside firing, in
a given 10 ms time frame.
- This is more complex than the stable state of a Hopfield Net.
In a synchronous Hopfield net, all neurons maintain the same
state (either on or off).
- Like Hopfield nets, there are Attractor Dynamics.
- A wide range of initial states (external stimuli) will lead
to the same final stable state.
- In other words, a large number of stimuli activate the
same short term memory.
- For example, a lot of horses activate the horse concept.
- Moreover, a lot of different visual inputs activate my CA
of my wife.
