Cultural Counterfactuals is a high-quality synthetic image dataset for measuring cultural biases in Large Vision-Language Models (LVLMs). It contains 59,827 images organized into 10,331 counterfactual sets across three cultural dimensions: religion, nationality, and socioeconomic status. Within each set, the same synthetic individual is depicted in multiple distinct cultural contexts (e.g., the same person standing in front of a Christian church, a… See the full description on the dataset page:
https://huggingface.co/datasets/thoughtworks/CulturalCounterfactuals.
