Last week in the journal club, we discussed two varieties of reinforcement learning algorithms: the Resorla-Wagner model and the temporal difference learning model. A key part of both models is representing the expected reward resulting from a given action. Is there any evidence for such representations in the brain? This Friday, November 9th, at Sensorimotor Journal Club, I will present an article showing that neurons in the basal ganglia appear to represent these values:
Samejima, K., Ueda, Y., Doya, K., & Kimura, M. (2005). Representation of action-specific reward values in the striatum. Science, 310(5752), 1337-1340. (link to pdf of article)
The estimation of the reward an action will yield is critical in decision-making. To elucidate the role of the basal ganglia in this process, we recorded striatal neurons of monkeys who chose between left and right handle turns, based on the estimated reward probabilities of the actions. During a delay period before the choices, the activity of more than one-third of striatal projection neurons was selective to the values of one of the two actions. Fewer neurons were tuned to relative values or action choice. These results suggest representation of action values in the striatum, which can guide action selection in the basal ganglia circuit.