Measuring and Mitigating Interference in Reinforcement Learning