Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

Picture for Tom Goldstein

The Intrinsic Dimension of Images and Its Impact on Learning


Apr 18, 2021
Phillip Pope, Chen Zhu, Ahmed Abdelkader, Micah Goldblum, Tom Goldstein

* To appear at ICLR 2021 (spotlight), 17 pages with appendix, 15 figures 

  Access Paper or Ask Questions

THAT: Two Head Adversarial Training for Improving Robustness at Scale


Mar 25, 2021
Zuxuan Wu, Tom Goldstein, Larry S. Davis, Ser-Nam Lim


  Access Paper or Ask Questions

Insta-RS: Instance-wise Randomized Smoothing for Improved Robustness and Accuracy


Mar 21, 2021
Chen Chen, Kezhi Kong, Peihong Yu, Juan Luque, Tom Goldstein, Furong Huang


  Access Paper or Ask Questions

Thinking Deeply with Recurrence: Generalizing from Easy to Hard Sequential Reasoning Problems


Mar 17, 2021
Avi Schwarzschild, Arjun Gupta, Micah Goldblum, Tom Goldstein


  Access Paper or Ask Questions

Improving Generalization of Transfer Learning Across Domains Using Spatio-Temporal Features in Autonomous Driving


Mar 15, 2021
Shivam Akhauri, Laura Zheng, Tom Goldstein, Ming Lin

* 6 pages, 3 figures, 8 tables 

  Access Paper or Ask Questions

Preventing Unauthorized Use of Proprietary Data: Poisoning for Secure Dataset Release


Mar 05, 2021
Liam Fowl, Ping-yeh Chiang, Micah Goldblum, Jonas Geiping, Arpit Bansal, Wojtek Czaja, Tom Goldstein


  Access Paper or Ask Questions

DP-InstaHide: Provably Defusing Poisoning and Backdoor Attacks with Differentially Private Data Augmentations


Mar 02, 2021
Eitan Borgnia, Jonas Geiping, Valeriia Cherepanova, Liam Fowl, Arjun Gupta, Amin Ghiasi, Furong Huang, Micah Goldblum, Tom Goldstein

* 11 pages, 5 figures 

  Access Paper or Ask Questions

What Doesn't Kill You Makes You Robust(er): Adversarial Training against Poisons and Backdoors


Feb 26, 2021
Jonas Geiping, Liam Fowl, Gowthami Somepalli, Micah Goldblum, Michael Moeller, Tom Goldstein

* 17 pages, 14 figures 

  Access Paper or Ask Questions

Improving Robustness of Learning-based Autonomous Steering Using Adversarial Images


Feb 26, 2021
Yu Shen, Laura Zheng, Manli Shu, Weizi Li, Tom Goldstein, Ming C. Lin


  Access Paper or Ask Questions

Center Smoothing for Certifiably Robust Vector-Valued Functions


Feb 19, 2021
Aounon Kumar, Tom Goldstein


  Access Paper or Ask Questions

GradInit: Learning to Initialize Neural Networks for Stable and Efficient Training


Feb 16, 2021
Chen Zhu, Renkun Ni, Zheng Xu, Kezhi Kong, W. Ronny Huang, Tom Goldstein


  Access Paper or Ask Questions

Technical Challenges for Training Fair Neural Networks


Feb 12, 2021
Valeriia Cherepanova, Vedant Nanda, Micah Goldblum, John P. Dickerson, Tom Goldstein


  Access Paper or Ask Questions

LowKey: Leveraging Adversarial Attacks to Protect Social Media Users from Facial Recognition


Jan 25, 2021
Valeriia Cherepanova, Micah Goldblum, Harrison Foley, Shiyuan Duan, John Dickerson, Gavin Taylor, Tom Goldstein

* Published as a conference paper at ICLR 2021 

  Access Paper or Ask Questions

Dataset Security for Machine Learning: Data Poisoning, Backdoor Attacks, and Defenses


Dec 30, 2020
Micah Goldblum, Dimitris Tsipras, Chulin Xie, Xinyun Chen, Avi Schwarzschild, Dawn Song, Aleksander Madry, Bo Li, Tom Goldstein


  Access Paper or Ask Questions

Data Security for Machine Learning: Data Poisoning, Backdoor Attacks, and Defenses


Dec 18, 2020
Micah Goldblum, Dimitris Tsipras, Chulin Xie, Xinyun Chen, Avi Schwarzschild, Dawn Song, Aleksander Madry, Bo Li, Tom Goldstein


  Access Paper or Ask Questions

Analyzing the Machine Learning Conference Review Process


Nov 26, 2020
David Tran, Alex Valtchanov, Keshav Ganapathy, Raymond Feng, Eric Slud, Micah Goldblum, Tom Goldstein

* NeurIPS Workshop on Navigating the Broader Impacts of AI Research. Full version at arXiv:2010.05137 

  Access Paper or Ask Questions

Strong Data Augmentation Sanitizes Poisoning and Backdoor Attacks Without an Accuracy Tradeoff


Nov 18, 2020
Eitan Borgnia, Valeriia Cherepanova, Liam Fowl, Amin Ghiasi, Jonas Geiping, Micah Goldblum, Tom Goldstein, Arjun Gupta

* Authors ordered alphabetically 

  Access Paper or Ask Questions

An Open Review of OpenReview: A Critical Analysis of the Machine Learning Conference Review Process


Oct 26, 2020
David Tran, Alex Valtchanov, Keshav Ganapathy, Raymond Feng, Eric Slud, Micah Goldblum, Tom Goldstein

* 19 pages, 6 Figures 

  Access Paper or Ask Questions

Are Adversarial Examples Created Equal? A Learnable Weighted Minimax Risk for Robustness under Non-uniform Attacks


Oct 24, 2020
Huimin Zeng, Chen Zhu, Tom Goldstein, Furong Huang


  Access Paper or Ask Questions

FLAG: Adversarial Data Augmentation for Graph Neural Networks


Oct 19, 2020
Kezhi Kong, Guohao Li, Mucong Ding, Zuxuan Wu, Chen Zhu, Bernard Ghanem, Gavin Taylor, Tom Goldstein


  Access Paper or Ask Questions

Towards Accurate Quantization and Pruning via Data-free Knowledge Transfer


Oct 14, 2020
Chen Zhu, Zheng Xu, Ali Shafahi, Manli Shu, Amin Ghiasi, Tom Goldstein


  Access Paper or Ask Questions

Data Augmentation for Meta-Learning


Oct 14, 2020
Renkun Ni, Micah Goldblum, Amr Sharaf, Kezhi Kong, Tom Goldstein


  Access Paper or Ask Questions

Random Network Distillation as a Diversity Metric for Both Image and Text Generation


Oct 13, 2020
Liam Fowl, Micah Goldblum, Arjun Gupta, Amr Sharaf, Tom Goldstein


  Access Paper or Ask Questions

ProportionNet: Balancing Fairness and Revenue for Auction Design with Deep Learning


Oct 13, 2020
Kevin Kuo, Anthony Ostuni, Elizabeth Horishny, Michael J. Curry, Samuel Dooley, Ping-yeh Chiang, Tom Goldstein, John P. Dickerson


  Access Paper or Ask Questions

Prepare for the Worst: Generalizing across Domain Shifts with Adversarial Batch Normalization


Sep 21, 2020
Manli Shu, Zuxuan Wu, Micah Goldblum, Tom Goldstein


  Access Paper or Ask Questions

Preparing for the Worst: Making Networks Less Brittle with Adversarial Batch Normalization


Sep 18, 2020
Manli Shu, Zuxuan Wu, Micah Goldblum, Tom Goldstein


  Access Paper or Ask Questions