RT Danny Driess

RT Danny Driess What happens when we train the largest vision-language model and add in robot experiences? The result is PaLM-E , a 562-billion parameter, general-purpose, embodied visual-language generalist - across robotics, vision, and language. Website: