CABot3
- The last of our original CABots followed the gross topology.
![](figGrossTop.GIF)
- It had about 100,000 neurons.
- Verb learning was off.
- It took natural language commands from a user, with an avatar
in the environment.
- The commands were parsed by a reasonable neuro-cognitive model of
parsing, binding with short-term potentiation.
- Planning was done with Maes nets.
- Vision used on-off off-on center surround detectors for the retina, and
angle and edge detectors for V1.
- There were a few types of objects it recognised.
- It explored the four room environment, and memorised the location
of the four objects (cognitive map).
- It could then go to the room before the striped pyramid.