How does the brain represent positive and negative feelings on a second-to-second basis? While prior work like PINES has identified generalizable neural substrates of affect, the temporal dimension remains underexplored. Using fMRI during a Monetary Incentive Delay task with temporally intensive affective probes, we examine whether spatially and temporally distinct circuits encode momentary affective experience, and how this may relate to reinforcement learning theories. Our findings suggest that decomposing affect into positive and negative arousal dimensions reveals dissociable neural substrates with distinct temporal profiles, offering a richer framework for linking brain activity to both subjective experience and reward-driven behavior.