Understanding the Generalization of In-Context Learning in Transformers: An Empirical Study

Open in new window