1 paper across 1 session
HoloV prunes 88.9% visual tokens in MLLMs while retaining 95.8% performance via spatial visual holistic context retention.
