GPT-4V

New ConTextual Dataset Evaluates Large Multimodal Models for Complex Visual Reasoning

Researchers from UCLA introduce a new dataset to benchmark large multimodal models' performance on context-sensitive text-rich vis