VideoSAVi: Self-Aligned Video Language Models without Human Supervision

VLM
Re-assign community