Divided control by past behavior, present stimuli, and future outcome value in a concurrent‐chains procedure | Synapse