AI RESEARCH

AutoDrive-R$^2$: Incentivizing Reasoning and Self-Reflection Capacity for VLA Model in Autonomous Driving

arXiv CS.CV

ArXi:2509.01944v3 Announce Type: replace-cross Vision-Language-Action (VLA) models in autonomous driving systems have recently nstrated transformative potential by integrating multimodal perception with decision-making capabilities. However, the interpretability and coherence of the decision process and the plausibility of action sequences remain largely underexplored.