Hugging Face· PlatformResearch paper

Introducing ConTextual: How well can your Multimodal model jointly reason over text and image in text-rich scenes?

Researchers from UCLA introduce ConTextual, a dataset and leaderboard for evaluating multimodal models on context-sensitive text-rich visual reasoning tasks. Initial experiments show that both proprietary and open-source models struggle on this benchmark compared to humans.

From the source

That’s why we (researchers from University of California Los Angeles) created ConTextual, a Context-sensitive Text-rich visuaL reasoning dataset for evaluating LMMs. We also released a leaderboard, so that the community can see for themselves which models are the best at this task.

huggingface.co

Announced

05 Mar 2024 00:00 UTC

Detected

03 Aug 2026 21:12 UTC (+881d after)