Figure 2.

Action and reward coding by striatal neurons. (A) Example striatal neuron active before movement (go) and silent before no-movement (no-go). Based on Schultz and Romo (1988), reproduced with permission. (B) Example striatal neurons coding reward. First row depicts a neuron with phasic active after juice reward delivery independent of the action to obtain reward. Second row depicts a neuron with tonic activity after juice reward delivery. Third row shows a neuron with tonic activity after no reward is delivered. Based on Hollerman et al. (1998), reproduced with permission. (C) Example caudate neuron coding the conjunction of action and reward. This neuron is active during the presentation of a cue indicating the saccade necessary to complete the trial if the trial will be rewarded (rewarded direction is highlighted by a bulls eye). R, right; U, up; L, left; D, down. Polar plots show the average response for each cue and direction. Based on Kawagoe et al. (1998), reproduced with permission. (D) (Top) Depiction of the probability of larger rewards associated with left or right actions on each condition block. Colored numbers refer to the probability associated with left-right actions. (Bottom) Example striatal neuron coding right action value. Based on Samejima et al. (2005), reproduced with permission.