Clarification on USMLE Questions

[hyesunyun]

Hello!

Thank you for making the evaluation datasets available.

In the README, it says that the USMLE questions are from "Automated Medical Question Answering: A Survey" - Rodríguez-Mier et al..

When I click on the link, I am directed to an article with a different name: "Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models." Is this the correct article? When I do a search on Google Scholar, I am not able to find an article named "Automated Medical Question Answering: A Survey" so wanted to make sure I am looking at the correct one.

I am able to find links to steps 1, 2, and 3 questions that were used to evaluate ChatGPT in the Supporting Information section. I believe these questions were filtered to exclude questions with visual assets. For creating the step_exams.jsonl file, I assume you only included questions with the MC-NJ tag (USMLE verbatim)?

Thanks!