HeiSD: Hybrid Speculative Decoding for Embodied Vision-Language-Action Models with Kinematic Awareness

作者：Zihao Zheng, Zhihao Mao, Sicheng Tian, Jiayu Chen, Maoliang Li, Xinhao Sun, Zhaobo Zhang, Xuanzhe Liu, Donggang Cao, Hong Mei, Xiang Chen · 单位：School of Computer Science, Peking University, School of Computer Science, China University of Geosciences (Wuhan), School of Artificial Intelligence, Beijing Normal University, School of EECS, Peking University · 会议/期刊：arXiv · 日期：2026-03-18 · 来源：Low-Level Learning-Based Action Modelling / Input Modelling / 2D Vision Language Action Models with Efficiency / Action Decoding