V-JEPA-LM: Video-Language Models That Read Dynamics Beyond Static, Text-Aligned Representations Paper • 2611.17831 • Published 3 days ago • 17