Topin, N.; Veloso, M. Generation of Policy-Level Explanations for Reinforcement Learning. AAAI 2019, 33, 2514-2521.