Datasets:

---

xytian1008

/

MUPO-Thinker-val1k

like 1

Hi! I'm Niels, part of the community science team at Hugging Face. I've updated this dataset card to include:

• A link to the associated paper: "All Roads Lead to Rome: Incentivizing Divergent Thinking in Vision-Language Models" (CVPR 2026).
• Links to the official GitHub repository and project page.
• The image-text-to-text task category in the metadata.
• A brief description of the dataset and its role in the MUPO framework.
• The BibTeX citation for the paper.