Our work on Bayesian parameter-efficient fine-tuning of vision foundation models appeared at NeurIPS 2024.