Text this: LLM-Guided Ensemble Learning for Contextual Bandits with Copula and Gaussian Process Models.