Reinforcement learning or active inference?

scientific article