For me, this is just as much about learning specific implementation of "correct reasoning" as it is about learning how the smartest humans work and what they update on. In retrospect, I probably should have clarified that more in my application. (Private message me if you want to see my application. I saved it on my local individual computer device.)