A-VL: Adaptive Attention for Large Vision-Language Models

VLM
Re-assign community