Dataset · Hugging Face

gengyuanmax/MosaicBench

Name: MosaicBench
Creator: gengyuanmax
License: other
Keywords: task_categories:visual-question-answering, language:en, license:other, size_categories:n<1K, format:json, modality:image, modality:text, library:datasets, library:pandas, library:polars, library:mlcroissant, arxiv:2609.39363, region:us

MosaicBench 560 questions across 28 task families, with 20 questions per family, in the original test split. Original prompts, ordered candidate answers and correct option labels are retained. No samples are replaced or silently dropped. Images and references 780 processed images are included. All 100 COCO-derived figures are reference-only: the manifest records original filenames, COCO IDs, original-image URLs and Flickr photo pages. These 100 figure references… See the full description on the dataset page: https://huggingface.co/datasets/gengyuanmax/MosaicBench.

View on Hugging FaceReport an issue

Author / organization	gengyuanmax
License (as declared)	other
Downloads	0 (as reported by Hugging Face)
Likes	0
Created	16 October 2026
Last updated at source	16 October 2026

Embed this entry

<iframe src="https://silytics.si/embed/item/93846db7-b731-4832-ae20-be15b448be21" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Connected in the knowledge graph

Rethinking Multi-Image Re-Representation in Multi-Image UnderstandingPaper · cites

Explore this entry in the graph

Related papers

Rethinking Multi-Image Re-Representation in Multi-Image UnderstandingPaper

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 16 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.