It is also reported that the activity of target neurons of dopamine signals in the striatum represent the reward values of action options ( 18, 19) and chosen actions ( 19, 20) during behavioral tasks in which experimental animals learn to choose options with a higher reward probability on a trial-and-error basis and to keep choosing the option even if those choices sometimes lead to no reward. The discounted value of a single reward that is expected after some delay is represented by neuronal activity in the cerebral cortex ( 11– 13) and dopamine neurons ( 14, 15) in rodents and primates, and, as shown with functional brain imaging, in the cerebral cortex and striatum in humans ( 16, 17). These findings demonstrate that dopamine neurons learn to encode the long-term value of multiple future rewards with distant rewards discounted.ĭopamine neurons convey the reward value signals of external events ( 2– 4), their expectation errors ( 2, 3, 5– 8), and signals of the motivational salience of external stimuli ( 5, 9, 10). The dopamine responses were quantitatively predicted by theoretical descriptions of the value function with time discounting in reinforcement learning. More specifically, when monkeys learned the multistep choice task over the course of several weeks, the responses of dopamine neurons encoded the sum of the immediate and expected multiple future rewards. In accordance with this behavioral observation, the dopamine responses to the start cues and reinforcer beeps reflected the expected values of the multiple future rewards and their errors, respectively. An analysis of anticipatory licking for reward water indicated that the monkeys did not anticipate an immediately expected reward in individual trials rather, they anticipated the sum of immediate and multiple future rewards. We recorded 185 dopamine neurons in three monkeys that performed a multistep choice task in which they explored a reward target among alternatives and then exploited that knowledge to receive one or two additional rewards by choosing the same target in a set of subsequent trials. Here, we address this experimentally untested issue. If they play a critical role in achieving specific distant goals, long-term future rewards should also be encoded as suggested in reinforcement learning theories. Midbrain dopamine neurons signal reward value, their prediction error, and the salience of events.
0 kommentar(er)
