Vision-Language Model Adaptation for Generalizable Autonomous Driving Systems. JDIAS 2026, 1 (3).