Experiment 1 Description
Axonal weights in an unaccessed network
- Using a 20 by 20 grid of CANT neurons, connected in
a toroidal topology
- Neurons are 20% inhibitory, 80% excitatory
- Run an untrained net with spontaneous activation only
- There are two conditions:
- strict Hebbian learning, constant weight change
- post-Hebbian learning based on existing weights
- Check the weights of all excitatory and inhibitory neurons
as the network evolves
