V-JEPA-LM: Video-Language Models That Read Dynamics Beyond Static, Text-Aligned Representationswith Keuntae Kim, Yong Suk Choi1 versionW_qy8e7kacv1 · current