LRM reasoning breaks down down past an unfamiliarity threshold, not "complexity" twitter.com 5 points by k1m 6 months ago · 0 comments Reader PiP Save No comments yet.