1 code implementation • NeurIPS 2021 • Omer Elkabetz, Nadav Cohen
The extent to which it represents gradient descent is an open question in the theory of deep learning.
Computational Efficiency Learning Theory +1