Search for a command to run...
HiF-VLA: Hindsight, Insight and Foresight through Motion Representation for Vision-Language-Action Models