TA-VLA: Elucidating the Design Space of Torque-aware Vision-Language-Action Models

Open in new window