Ma, Y. J.; Shen, A.; Bastani, O.; Dinesh, J. Conservative and Adaptive Penalty for Model-Based Safe Reinforcement Learning. AAAI 2022, 36, 5404-5412.