Reward predication error (RPE)

Reward predication error (RPE) : Dopaminergic activity encodes the anticipation of an external stimulus or reward, that is, the deviation or error between the reward actually received and the predicted timing and size of the reward. There are three conditions: 1. No prediction, reward-positive error - dopamine neurons show positive signals (spike production increases), and dopamine neurons are activated by the unexpected reward stimulus; 2. Prediction, reward - no error - no dopaminergic firing (no change in spike generation); The conditioned stimulus predicted reward, and reward occurred according to the prediction, so there was no error in the prediction of reward. Dopamine neurons were predicted to be activated by the stimulus of that reward, but not by the reward itself. 3. Forced loss, prediction, no reward -- negative error -- negative signal (reduced spike production); The conditioned stimulus predicted reward, which did not occur. The activity of dopamine neurons is inhibited. In the absence of any intervening stimulus, depression will occur after the conditioned stimulus.