GRU 1.5

Software / AppMentioned in 1 video

A vision-language action model that achieved the best overall performance in benchmarking humanoid manipulation tasks.