VL-JEPA: Joint Embedding Predictive Architecture for Vision-language Paper • 2512.10942 • Published 13 days ago • 3
