Return to Article Details Mechanistic Interpretability of In-Context Learning in Transformers Download Download PDF