Mildly Overparametrized Neural Nets can Memorize Training Data EfficientlyPublished in Preprint, 2019Share on Twitter Facebook LinkedIn Previous Next