Robot-R1: Reinforcement Learning for Enhanced Embodied Reasoning in Robotics | Read Paper on Bytez