S
Shibani Santurkar
Researcher at Massachusetts Institute of Technology
Publications - 50
Citations - 5767
Shibani Santurkar is an academic researcher from Massachusetts Institute of Technology. The author has contributed to research in topics: Robustness (computer science) & Computer science. The author has an hindex of 22, co-authored 44 publications receiving 4005 citations. Previous affiliations of Shibani Santurkar include Indian Institute of Technology Bombay.
Papers
More filters
Proceedings Article
Robustness May Be at Odds with Accuracy
TL;DR: It is shown that there may exist an inherent tension between the goal of adversarial robustness and that of standard generalization, and it is argued that this phenomenon is a consequence of robust classifiers learning fundamentally different feature representations than standard classifiers.
Proceedings Article
Adversarial Examples Are Not Bugs, They Are Features
Andrew Ilyas,Shibani Santurkar,Dimitris Tsipras,Logan Engstrom,Brandon Tran,Aleksander Madry +5 more
TL;DR: It is demonstrated that adversarial examples can be directly attributed to the presence of non-robust features: features derived from patterns in the data distribution that are highly predictive, yet brittle and incomprehensible to humans.
Proceedings Article
How does batch normalization help optimization
TL;DR: In this article, the authors uncover a more fundamental impact of batch normalization on the training process: it makes the optimization landscape significantly smoother, which induces a more predictive and stable behavior of the gradients, allowing for faster training.
Proceedings Article
Adversarially Robust Generalization Requires More Data
TL;DR: In this paper, the authors study adversarially robust learning from the viewpoint of generalization and show that the sample complexity of robust learning can be significantly larger than that of "standard" learning.