1 paper
Andy Gray
A central question in the debate over large language models is whether transformers can learn rules they have never seen, or whether they can only interpolate: predict new cases fr…