Dopamine potently regulates forebrain physiology that underlies reward learning and behavioral flexibility. A detailed accounting of how cellular and circuit interactions come to impact reinforcement learning continues to be refined at multiple levels of analysis. This review provides a brief summary of our current understanding of quantitative decision variables relayed by dopamine dynamics in striatal targets, circuit substrates that support specific dopaminergic computations, and how different behavioral task-demands leverage these specializations to afford adaptive behaviors. Moreover, we motivate the revision of a longstanding hypothesis for globally broadcast dopamine prediction error signals, and instead, make the case for regionally specialized forebrain dopamine dynamics tailored to local computational needs.
Bibliographical noteFunding Information:
This work was supported by the Hannah Gray Fellowship from Howard Hughes Medical Institute .
© 2021 Elsevier Ltd