cs.RONov 9, 2026

AutoVLA-D: Learn a Trajectory-Geometric Condition from Reasoning for Autonomous Driving VLA

Authors: Yuguang Yang, Zhewen Tan, Canyu Chen, Linlin Yang, Cheng Chi, Chunyang Liu, Kehua Sheng, Jinyu Yang, 

Organizations: School of Electronic and Information Engineering, Beihang University · Institute for AI Industry Research (AIR), Tsinghua University · School of Computer Science and Engineering, Beihang University · National College for Excellent Engineers, Beihang University · 

Abstract

Autoregressive VLA planning usually predicts continuous trajectories as the final output, but driving instructions often describe actions rather than ego states, which weakens direct conditioning on the final continuous trajectory. We address this gap with TrajCond, which introduces a discrete trajectory-geometric condition for VLA planning. Specifically, we represent TrajCond as a 1D-Brohan differential action trajectory and learn it with a two-stage training framework before final continuous trajectory prediction. With the simultaneous autoregressive generation of discrete and numeric trajectories, we name our method AutoVLA-D (Double). Experiments on NAVSIM navtest, nuScenes, and NeuroNCAP show that discrete trajectories are weaker as direct outputs but more effective as TrajCond, consistently improving final continuous trajectory prediction across closed-loop and open-loop driving benchmarks. Code is available at https://github.com/ShivaTeam/AutoVLA-D.

Figures & tables

Appendix figures & tables12 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Planning in the Backbone: DiffAdapterVLA for Native Continuous Trajectory Generation with Driving VLMs

    Sep 14, 2026Changxin Lu, Xiaoliang Meng, Yu Wu +5Trajectory GenerationVLMs for Autonomous Driving

  2. AnchorVLA: Bridging Discrete Decisions and Continuous Trajectories for Vision-Language-Action Planning

    Jul 3, 2026Qi Liu, Yabei Li, Hongsong Wang +2Hierarchical PlanningVision-Language-Action Models

  3. Teaching Vision-Language-Action Models What to See and Where to Look

    Jul 2, 2026Yuguang Yang, Canyu Chen, Zhewen Tan +10End-to-End Autonomous DrivingVLM Distillation