---
license: cc-by-4.0
---
**Real-world data in MultipanelVQA**

Paper: Muffin or Chihuahua? Challenging Large Vision-Language Models with Multipanel VQA [(arXiv)](https://arxiv.org/abs/2401.15847)

Website: [https://sites.google.com/view/multipanelvqa/home](https://sites.google.com/view/multipanelvqa/home)

MultipanelVQA includes both real-world subset and [synthetic subset](https://huggingface.co/datasets/yfan1997/MultipanelVQA_synthetic).