R
Rebecca Roelofs
Researcher at Google
Publications - 27
Citations - 2118
Rebecca Roelofs is an academic researcher from Google. The author has contributed to research in topics: Computer science & Overfitting. The author has an hindex of 12, co-authored 22 publications receiving 1411 citations. Previous affiliations of Rebecca Roelofs include Swarthmore College & University of California, Berkeley.
Papers
More filters
Proceedings Article
The Marginal Value of Adaptive Gradient Methods in Machine Learning
TL;DR: It is observed that the solutions found by adaptive methods generalize worse (often significantly worse) than SGD, even when these solutions have better training performance, suggesting that practitioners should reconsider the use of adaptive methods to train neural networks.
Posted Content
Do CIFAR-10 Classifiers Generalize to CIFAR-10?
TL;DR: This work measures the accuracy of CIFAR-10 classifiers by creating a new test set of truly unseen images and finds a large drop in accuracy for a broad range of deep learning models.
Posted Content
The Marginal Value of Adaptive Gradient Methods in Machine Learning
TL;DR: This article showed that adaptive methods often find drastically different solutions than gradient descent or stochastic gradient descent (SGD) for simple overparameterized problems, and that the solutions found by adaptive methods generalize worse (often significantly worse) than SGD, even when these solutions have better training performance.
Posted Content
Do ImageNet Classifiers Generalize to ImageNet
TL;DR: This article showed that the accuracy drops are not caused by adaptivity, but by the models' inability to generalize to slightly "harder" images than those found in the original test sets.
Proceedings Article
Do ImageNet Classifiers Generalize to ImageNet
TL;DR: This article showed that the accuracy drops are not caused by adaptivity, but by the models' inability to generalize to slightly "harder" images than those found in the original test sets.