Arthur Westbrook, “Facilitating Zero Shot Decision Generalization through Conservative Offline Reinforcement Learning and Semantic Policy Pre training with Large Language Models”, IJAIR, vol. 1, no. 2, May 2026.