C. Yang, M. Littman, and M. Carbin, “Computably Continuous Reinforcement-Learning Objectives Are PAC-Learnable”, AAAI, vol. 37, no. 9, pp. 10729-10736, Jun. 2023.