Lonh, N. 2022. Uncovering Text-to-Image Drift Models’ Separation Ability. International Transactions in Machine Learning. 4, 4 (Dec. 2022).