1 paper · 1 filter
Bosung Kim, Ruiyi Wang, David Acuna +5
Scaling robot policy learning is bottlenecked by the cost of collecting demonstrations, while language annotations for existing demonstrations are comparatively cheap. We study lan…