3 ms·RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control1 points by gavi 3y ago