Skip to yearly menu bar Skip to main content


Findings Poster Sat, Jun 6, 2026 • 6:30 AM – 8:00 AM PDT ExHall A 278

PrismPrune: Decoupling Saliency and Diversity in Attention for Efficient Visual Token Pruning in VLMs

Ziniu Liu, Shuheng Zhou, Mingqing Liu, Hao Deng, Huijia Zhu
Keywords: Multimodal Learning

Log in and register to view live content