On Robustness and Chain-of-Thought Consistency of RL-Finetuned VLMs - Apple Machine Learning Research

On Robustness and Chain-of-Thought Consistency of RL-Finetuned VLMs Apple Machine Learning Research