Submitted by Siteng Huang 11 HiF-VLA: Hindsight, Insight and Foresight through Motion Representation for Vision-Language-Action Models Westlake Robotics Technology (Hangzhou) Co., Ltd. 35 2