Fetching the paper…

ConTextual: Evaluating Context-Sensitive Text-Rich Visual Reasoning in Large Multimodal Models · Around