S. Joshi, J. Zhang, and E. Bareinboim, “Towards Safe Policy Learning under Partial Identifiability: A Causal Approach”, AAAI, vol. 38, no. 12, pp. 13004–13012, Mar. 2024.