2024 ICML ICML 2024

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