Identifiability in Inverse Reinforcement Learning: Supplementary Material