S. Cong, C. Yu, and X. Lan, “CATAL: Causally Disentangled Task Representation Learning for Offline Meta-Reinforcement Learning”, AAAI, vol. 40, no. 25, pp. 20634–20641, Mar. 2026.