Fast and far generalization from sparse data Lei Yuan Indiana University Much learning is slow, incremental, and limited in generalizability. However, there are clear cases of fast learning with far generalization from quite limited input …