Talk 2

I gave a talk about our paper ​What learning algorithm is in-context learning? Investigations with linear models at MIT Mechanistic Interpretability Conference.