VL-Rethinker: Incentivizing Self-Reflection of Vision-Language Models with Reinforcement Learning Paper • 2504.08837 • Published 13 days ago • 42