Using Natural Language and Program Abstractions to Instill Human Inductive Biases in Machines

Kumar, Sreejan, Correa, Carlos G., Dasgupta, Ishita · arXiv · 2022