How the Level of Reward Awareness Changes the Computational and Electrophysiological Signatures of Reinforcement Learning