Dopamine is not the whole story for reward! We found a VTA GABA projection to ventral pallidum that encodes reward and does not change with experience. Is this the information needed for proper prediction error calculation?