acruz/folktexts

Dataset Card for folktexts Folktexts is a suite of Q&A datasets with natural outcome uncertainty, aimed at evaluating LLMs' calibration on unrealizable tasks. The folktexts datasets are derived from US Census data products. Namely, the datasets made available here are derived from the 2018 Public Use Microdata Sample (PUMS). Individual features are mapped to natural text using the respective codebook. Each task relates to predicting different individual characteristics (e.g.… See the full description on the dataset page: https://huggingface.co/datasets/acruz/folktexts.