Curious about how text-to-image diffusion models behave when prompted with psychiatric diagnosis inputs? See examples and reflections in my first 1st author paper, now available in @BMJMentalHealth (free): mentalhealth.bmj.com/content/27/1...