In a vast simplification, the mid-brain sends reward signals (for example through dopaminergic neurons) that tell the rest of the brain whether it succeeded at fulfilling the needs of the organism. If there are certain activities in the brain that lead to a successful action, then these activities should have a higher probability of occurring again (since that is useful for achieving the organism's needs in future).
To increase the probability of the occurrence of an activation pattern as a response to a certain history of sensory input, the neurons that were involved to produce this pattern will need to adapt their synapses such that they become more sensitive to this input.
I'm wondering, how are the reward signals thought to affect the neurons that were involved in the previous actions of the organism at the synaptic level?
