Response profiles of same three neurons in the top (fourth) layer of VisNet after training.

<p>Response profiles of the same three neurons from Fig. 3 after training on the images shown in Fig. 1. Conventions as in Fig. 3. It can be seen that each of the three cells responds selectively to just one of the hand-object configurations, and responds to that configuration over all five retinal positions shown along the abscissae. Moreover, each of the three hand-object configurations is represented by one of the cells.