[2603.01581] KERV: Kinematic-Rectified Speculative Decoding for Embodied VLA Models

Computer Science > Robotics arXiv:2603.01581 (cs) [Submitted on 2 Mar 2026] Title:KERV: Kinematic-Rectified Speculative Decoding for Embodied VLA Models Authors:Zihao Zheng, Zhihao Mao, Maoliang Li, Jiayu Chen, Xinhao Sun, Zhaobo Zhang, Donggang Cao, Hong Mei, Xiang Chen View a PDF of the paper titled KERV: Kinematic-Rectified Speculative Decoding for Embodied VLA Models, by Zihao Zheng and 7 other authors View PDF HTML (experimental) Abstract:Vision-Language-Action (VLA) models build a token-domain robot control paradigm, yet suffer from low speed. Speculative Decoding (SD) is an optimization strategy that can boost inference speed. Two key issues emerge when integrating VLA and SD: first, SD relies on re-inference to address token errors, which is computationally expensive; second, to mitigate token errors, the acceptance threshold in SD requires careful adjustment. Existing works fail to address the above two issues effectively. Meanwhile, as the bridge between AI and the physical world, existing embodied intelligence has overlooked the application of robotic kinematics. To address these issues, we innovatively combine token-domain VLA models with kinematic-domain prediction for SD, proposing a kinematic-rectified SD framework named KERV. We employ a kinematics-based Kalman Filter to predict actions and compensate for SD errors, avoiding costly re-inference. Moreover, we design a kinematics-based adjustment strategy to dynamically rectify the acceptance threshold, addre...