Electrophysiological Correlates of Reward Processing in Dopamine Neurons

TLDR: In this paper, the dopamine reward-prediction error signal is used for economic choices that maximize utility, and the dopamine signal fits well into formal competitive decision models, whereby it codes the output variable (chosen value) suitable for updating or immediately influencing main input variables (object value and action value).

Abstract: Studies have identified three novel properties of the dopamine reward-prediction error signal First, the dopamine response reports initially and unselectively many salient, potentially rewarding events and subsequently processes more specifically the reward-prediction error This two-component structure restricts the earlier claimed salience coding to the initial component and explains aversive activations by physical impact rather than punishment Second, the dopamine prediction error signal reflects subjective reward value and, more stringently, formal economic utility A dopamine utility prediction error signal would be particularly useful for economic choices that maximize utility Third, the dopamine signal fits well into formal competitive decision models, whereby it codes the output variable (chosen value) suitable for updating or immediately influencing main input variables (object value and action value) With these properties, the dopamine utility prediction error signal bridges the gap between animal learning theory (prediction error) and economic decision theory (utility)