DriveReferee: Geometric Safety Verdicts Need Not Be Learned for Driving World-Action Models

arXiv:2609.22762 · cs.CV · Submitted 2026-09-19 · Read on arXiv

cs.CV

Submitted: 2026-09-19

Updated: 2026-09-19

Terminology

• Verifier — 21× in this paper · explained in 2 episodes
• Distillation — 13× in this paper · explained in 2 episodes
• Verification — 5× in this paper · explained in 2 episodes
• World Model — 5× in this paper · explained in 7 episodes
• Fidelity — 4× in this paper · explained in 2 episodes
• Ablation — 2× in this paper · explained in 1 episode
• Benchmark — 2× in this paper · explained in 2 episodes
• No At-Fault Collision — 2× in this paper
• Toyota Research Institute — 2× in this paper
• Action Flow — 1× in this paper · explained in 1 episode
• Adaptive Expert Routing — 1× in this paper · explained in 1 episode
• Adaptive Reasoning — 1× in this paper · explained in 2 episodes
• Alignment — 1× in this paper · explained in 1 episode
• Benchmarking — 1× in this paper · explained in 4 episodes
• Calibration — 1× in this paper · explained in 7 episodes
• Diffusion Model — 1× in this paper · explained in 8 episodes
• Drivable Area Compliance — 1× in this paper
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• Flow Matching — 1× in this paper · explained in 8 episodes
• Multimodal Dataset — 1× in this paper · explained in 2 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Transformer — 1× in this paper · explained in 2 episodes

Sources

• UNIVERSE: Unified Video Action Models for Autonomous Driving with Flexible Mask-Modulated Modality Generation
• DriveVA: Video Action Models are Zero-Shot Drivers
• DriveWAM: Video Generative Priors Enable Scalable World-Action Modeling for Autonomous Driving
• DriveDreamer-Policy: A Geometry-Grounded World-Action Model for Unified Generation and Planning
• Metis: A Generalizable and Efficient World-Action Model for Autonomous Driving and Urban Navigation
• Hydra-MDP: End-to-end Multimodal Planning with Multi-target Hydra-Distillation
• Hydra-MDP++: Advancing End-to-End Driving via Expert-Guided Hydra-Distillation
• Distill to Think, Foresee to Act: Cognitive-Physical Reinforcement Learning for Autonomous Driving
• Beyond Imitation: Learning Safe End-to-End Autonomous Driving from Hard Negatives
• PerceptDrive: Perception Prior World-Action Modeling with Adaptive Expert Routing for End-to-End Autonomous Driving
• DriveVer: Lightweight Trajectory Evaluator as Test-Time Verifier for Autonomous Driving
• Test-Time Trajectory Optimization for Autonomous Driving
• EvaDrive: Evolutionary Adversarial Policy Optimization for End-to-End Autonomous Driving
• MindDrive: An All-in-One Framework Bridging World Models and Vision-Language Model for End-to-End Autonomous Driving
• On a Formal Model of Safe and Scalable Self-driving Cars
• Cosmos 3: Omnimodal World Models for Physical AI
• NuPlan: A closed-loop ML-based planning benchmark for autonomous vehicles
• SparseDriveV2: Scoring is All You Need for End-to-End Autonomous Driving
• ForgeDrive: Bidirectional Cross-Conditioning for Unified Visual-Action Generation in Autonomous Driving

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models