RESEARCH

Our Research

Dopamine neurons and reward coding

Reward predictions: Dopamine neurons do not respond to reward per se, they respond to the difference between reward and predicted reward. By coding for this difference then, they should process and reflect reward (minuend) AND prediction (subtrahend). We know quite a bit about how dopamine neurons process reward value. They integrate different reward attributes and code for a highly-specified form of subjective reward value, economic utility, that places the value of different rewards on a common scale for easy comparison. Well-defined and easily measured utility functions, therefore, provide a rigorous account of the dopaminergic minuend. However, we know less about the dopaminergic subtrahend: reward predictions. The classic model for dopamine activity, the TD model, predicts the time-discounted expected value of future rewards. Although this quantity is reflected in dopamine signals, reward predictions probably go well beyond simple first-order reward statistics. We are interested in learning about the content of neural predictions.

Lak, A., Stauffer, W.R., Schultz, W. Dopamine neurons learn relative chosen value from probabilistic rewards. eLife (2016) 5: e18044

Stauffer, W.R, Lak, A., Schultz, W. Dopamine reward prediction error responses reflect marginal utility. Current Biology (2014) 24: 2491-2500

Lak, A., Stauffer, W.R, Schultz, W. Dopamine prediction error responses integrate subjective value from different reward dimensions. PNAS (2014) 111(6): 2343-2348

Monkey optogenetics

Optogenetics is one of the most exciting technologies around. However, the application of cell type-specific manipulations such as optogenetics to nonhuman primates is limited by a dearth of small, cell type-specific promoters to direct expression. We are interested in the cell types that comprise different parts of the monkey brain, and we are investigating promoters to allow us to access these cells.

Stauffer, W.R., Lak, A.,Yang, A., Borel, M., Paulsen, O., Boyden, E.S., Schultz, W. Dopamine neuron-specific optogenetic stimulation in Rhesus macaques. Cell (2016) 166(6): 1561-1571

Galvan A, Stauffer W.R., Acker L, El-Shamayleh Y, Inoue KI, Ohayon S, Schmid MC. Nonhuman Primate Optogenetics: Recent Advances and Future Directions. The Journal of Neuroscience (2017) 37(45):10894-10903.

Economic decision making

Uncertainty: Although we choose to get the best rewards, we can never predict the future. Therefore, uncertainty is a critical component of values and decisions. We are interested in different ‘forms’ of uncertainty, including risk and ambiguity, and how these influence value and decision making. We use decision-making paradigms with Rhesus monkeys to uncover attitudes about uncertainty, and we use electrophysiological recordings from dopamine neurons to see how ambiguity effects the neural coding of reward.

Genest, W., Stauffer, W.R., Schultz, W. Utility functions predict variance and skewness risk preferences in monkeys. PNAS (2016) 113(30): 8402-8407

Stauffer, W.R., Lak, A., Bossaerts, P., Schultz, W. Economic choices reveal probability distortion in rhesus macaques. Journal of Neuroscience (2015) 35(7): 3146-3154

Stauffer, W.R, Lak, A., Schultz, W. Dopamine reward prediction error responses reflect marginal utility. Current Biology (2014) 24: 2491-2500

Deliberation and optimality: Individuals often face complex choices where value is not easily realized and the goal is to select the optimal rewards. The neurocomputational mechanisms of deliberation likely involve reward and memory structures including the frontal cortex, basal ganglia (including dopamine neurons), and hippocampus. The goals of this research are to formalize the investigation of value deliberation using mathematical optimization theory, and to uncover the neuronal algorithms and implementation of optimization over value. We are developing new behavioral paradigms and employing optogenetics and electrophysiology to investigate the role of specific cell types.