Grid2Matrix: Revealing Digital Agnosia in Vision-Language Models Paper • 2604.09687 • Published 21 days ago • 8