Personalized dermatologic communication with large language models on high‐stakes personas

The datasets consist of anonymized clinical vignettes developed by the authors and outputs generated by large language models. Summary results are contained within this article. The full set of model-generated outputs can be made available from the corresponding author upon reasonable request.