From Segments to Scenes: Temporal Understanding in Autonomous Driving via Vision-Language Model Paper • 2512.05277 • Published Dec 4, 2025 • 6 • 2