Paper

V-JEPA-LM: Video-Language Models That Read Dynamics Beyond Static, Text-Aligned Representations · ByteTrail