cs.RONov 18, 2026

Efficient World-Model Planning via Physics-Guided Search Distillation

Authors: Hongyu Zhang, Xiaoshuai Hao, Heye Huang, Qingwen Meng, Wenjun Shi, Jianao Xu, Xin Hu, Xuanru Jin, 

Organizations: Yanshan University · Xiaomi Corporation · KAIST · Tsinghua University · 

Abstract

World models support autonomous driving by evaluating candidate trajectories through predicted future scene evolution. However, accurate future prediction does not ensure road-compliant execution; when all candidates violate road boundaries, re-ranking cannot select a feasible trajectory. We propose Physics-Guided Search Distillation (PGSD), a single-view planning framework that combines future representations and execution feedback for verified beyond-pool repair through three components: (1) Execution-Aware Selection checks ego-vehicle road compliance under simulated execution and selects the highest-scoring feasible trajectory; (2) a Clearance-Aware Search Teacher generates beyond-pool corrections and selects demonstrations with larger minimum road-boundary clearance during execution to buffer imitation errors; and (3) a Future-Conditioned Repair Student predicts one correction from rewarder features and execution feedback for single-rollout verification. The pretrained planner and rewarder remain frozen; no future scenes are generated online. On NAVSIM v1, single-view PGSD achieves 90.97 PDMS, exceeding all compared prior single-view planners and retaining 73.2% of the search teacher's PDMS gain over execution-aware selection. On NAVSIM v2 navhard, it reaches 38.96 EPDMS, improving on WorldDrive by 11.24 points (40.5%). The student replaces up to ten search rollouts with a single verification rollout per repair trigger. HUGSIM Hard examples provide qualitative evidence of zero-shot transfer.

Figures & tables

Appendix figures & tables4 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. World4Scorer: Outcome-Grounded World Modeling for Autonomous Driving

    Sep 29, 2026Jieyuan Pei, Meiyi Lu, Sining Ang +12World Model-Based PlanningWorld Models for Autonomous Driving

  2. DriveVA: Video Action Models are Zero-Shot Drivers

    Apr 5, 2026Mengmeng Liu, Diankun Zhang, Jiuming Liu +7Action-Conditioned Video GenerationWorld Models

  3. PlanWAM: Planning-Shaped Future Representations for End-to-End Autonomous Driving

    Oct 8, 2026Jinchang Xu, Hongda Yu, Fengwei Dong +9End-to-End Autonomous DrivingWorld Model-Based Planning