Picture for Shansong Yang

Shansong Yang

GRIP-VLM: Group-Relative Importance Pruning for Efficient Vision-Language Models

Add code
May 13, 2026
Viaarxiv icon