Paper

Measuring the Reasoning-Action Coupling in Driving Vision-Language-Action Models through Counterfactuals · ByteTrail