Fetching the paper…

Visionary-R1: Mitigating Shortcuts in Visual Reasoning with Reinforcement Learning · Around