Looking Inward: Language Models Can Learn About Themselves by Introspection Paper • 2410.13787 • Published Oct 17 • 5