This is really cool/necessary! First two indications I looked up had the wrong prevalence labels, though. Both diseases are more common than the assigned rare/ultra-rare labels. They were originally thought to be small populations, but higher prevalence numbers were published within the last few years.

@draparente Thanks! Do you mind sharing which ones? Agree that anchoring to highly visible/indexed sources is a problem with LLMs. Part of why we wanted to generate and share this is to identify areas where LLMs make mistakes

@Atelfo btw - love what you guys are doing! Wish you guys had individual accounts :)

@draparente We do! :) Thanks for the extra detail on the indications too, we'll look into it