Vidyaaranya Macha

Robot Learning with Super-Linear Scaling

Scaling robot learning requires data collection pipelines that scale favorably with human effort. In this work, we propose