Mental Health Stigma Benchmark Dataset
A theory-grounded, human-annotated benchmark of 470 news and social media excerpts for evaluating the detection and classification of mental health stigma in online communication.
Oct 10, 2026
Automatic Evaluation of Mental Health Stigma in Online Communication
We introduce a theory-grounded, human-annotated benchmark for evaluating mental health stigma across six conditions and 28 stigma categories, showing that LLMs frequently overpredict stigma and that conventional sentiment, toxicity, and hate-speech classifiers are insufficient.
Oct 2, 2026