Other Components of the Agent

The other two components of the agent are a motor area which sends commands to the agent.
Roman is working on this, and should have it done shortly.
The second component is an Activation Net which can be used for goals and decision making.
Our RA, Yulie Fan, is working on a symbolic version of this net.
In the long run (hopefully by Feb.) we hope to implement this in fLIF neurons.
Thus we're hoping to have a version 0.5 of the agent done by December.
Aside from the Activation net, it will be implemented entirely in fLIF neurons.
It should be able to handle simple commands like Move left.
It should also be able to resolve more complex commands like Turn toward the Pyramid.
It should be able to resolve some referents like Move towards it. when a Pyramid is present. This is the start of symbol grounding.

Other Components of the Agent (version 0.5)