A large-scale benchmark for few-shot program induction and synthesis

    Ferran Alet * 1 Javier Lopez-Contreras * 1 James Koppel 1 Maxwell Nye 1 Armando Solar-Lezama 1
            Tomás Lozano-Pérez 1 Leslie Pack Kaelbling 1 Joshua B. Tenenbaum 1

             Abstract
   A landmark challenge for AI is to learn fexible,
   powerful representations from small numbers of
   examples. On an important class of tasks, hy-
   potheses in the form of programs provide ex-
   treme generalization capabi ...