The CABots
- We've been following this framework for over a decade now,
and have had success with theoretical work (e.g. spontaneous activity,
learning rules and variable binding) and tasks (e.g. categorisation,
hierarchical categorisation, and rule implementation).
- The last four years we've concentrated on Cell Assembly roBots
(CABots).
- There have been three CABot1, CABot2, and the just completed
CABot3.
- They've all been based on our fatiguing leaky integrate and
fire (fLIF) neurons.
- They all view the environment, take commands from the user,
use plans, and act.
- CABot2 includes goal-action pair learning and a neural-cognitive model
of parsing.
- CABot3 integrates bump sensors, visual grating cells for texture
recognition, and spatial cognitive mapping.
- They run on a PC (and you can download them from
www.cwa.mdx.ac.uk/CABot/CANT.html).
- CABot3 is the Proto-Neural Cognitive Architecture.
