Datasetpublic

suparnojit/GenderVL-Bench

GenderVL-Bench GenderVL-Bench is a compact vision-language benchmark for evaluating how Vision-Language Models (VLMs) interpret gender-related representations across different occupations. Dataset 108 images 12 occupations 9 images per occupation Format: JPEG / ImageFolder Split: train Usage from datasets import load_dataset dataset = load_dataset("suparnojit/GenderVL-Bench") Citation @dataset{sarkar2026gendervlbench, author… See the full description on the dataset page: https://huggingface.co/datasets/suparnojit/GenderVL-Bench.

sourceHugging Faceupdated 1mo agoView on Hugging Face

1likes598downloads

Dataset Card

GenderVL-Bench

GenderVL-Bench is a compact vision-language benchmark for evaluating how Vision-Language Models (VLMs) interpret gender-related representations across different occupations.

Dataset

—108 images
—12 occupations
—9 images per occupation
—Format: JPEG / ImageFolder
—Split: train

Usage

python

from datasets import load_dataset

dataset = load_dataset("suparnojit/GenderVL-Bench")

Citation

bibtex

@dataset{sarkar2026gendervlbench,
  author    = {Sarkar, Suparnojit},
  title     = {GenderVL-Bench},
  year      = {2026},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/datasets/suparnojit/GenderVL-Bench}
}