[1]
E. D. Langlois and T. Everitt, “How RL Agents Behave When Their Actions Are Modified”, AAAI, vol. 35, no. 13, pp. 11586-11594, May 2021.