AnchorRefine: Synergy-Manipulation Based on Trajectory Anchor and Residual Refinement for Vision-Language-Action Models

作者：Tingzheng Jia · 单位：Beijing University of Technology, Beihang University, Nanjing University of Science and · 会议/期刊：arXiv · 日期：2026-04-20 · 来源：Low-Level Learning-Based Action Modelling / Input Modelling / 2D Vision Language Action Models with Model-agnostic Strategies