Figure 4.

Illustration of the “bias zone” effect. In each graph, the target was first hit N times (labeled at the top of the graph). Then, learning was turned off and movement for each possible value of G_exp was evaluated. Each dot represents the spatial location corresponding to G_exp. Large green dots represent locations of G_exp that resulted in movements that hit the location corresponding to G_exp. Small red dots represent locations of G_exp that, because of biasing implemented by weights onto D1 and D2, resulted in movements that hit the target (represented by the red circle in the lower right).