CoPhy distills VLM knowledge into a BEV encoder and uses an action-conditioned auto-regressive BEV world model inside GRPO with dual physical-cognitive rewards to reach SOTA on NAVSIM v1/v2 while adding language-based intent control.
Drivedpo: Policy learning via safety dpo for end-to-end autonomous driving.NeurIPS, 2025
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Distill to Think, Foresee to Act: Cognitive-Physical Reinforcement Learning for Autonomous Driving
CoPhy distills VLM knowledge into a BEV encoder and uses an action-conditioned auto-regressive BEV world model inside GRPO with dual physical-cognitive rewards to reach SOTA on NAVSIM v1/v2 while adding language-based intent control.