VisOnlyQA: Large Vision Language Models Still Struggle with Visual Perception of Geometric Information
Paper
•
2412.00947
•
Published
•
7
VisOnlyQA
Dataset for evaluating the visual perception capabilities of LVLMs.