Fetching the paper…

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