I am a postdoc at MIT with Aleksander Madry. Previously, I worked with Zico Kolter as a PhD student in the Machine Learning Department at Carnegie Mellon University, where I wrote my dissertation on robustness of deep networks to adversarial examples, which you can find here. My research interests touch on machine learning, robust optimization, and verification.

I am currently organizing a workshop on Robust and Reliable Machine Learning in the Real World for ICLR 2021

Publications

• Learning perturbation sets for robust machine learning
  Eric Wong, J. Zico Kolter
  In proceedings of the International Conference on Learning Representations (ICLR), 2021
  source code on Github Blog post
• Overfitting in adversarially robust deep learning
  Leslie Rice*, Eric Wong*, J. Zico Kolter
  In Proceedings of the International Conference on Machine learning (ICML), 2020
  source code on Github
• Adversarial robustness against the union of multiple perturbation models
  Pratyush Maini, Eric Wong, J. Zico Kolter
  In Proceedings of the International Conference on Machine learning (ICML), 2020
  source code on Github
• Neural network virtual sensors for fuel injection quantities with provable performance specifications
  Eric Wong, Tim Schneider, Joerg Schmitt, Frank R. Schmidt, J. Zico Kolter
  In Proceedings of the IEEE Intelligent Vehicles Syimposium (IV), 2020
  
• Fast is better than free: revisiting adversarial training
  Eric Wong*, Leslie Rice*, J. Zico Kolter
  In Proceedings of the International Conference on Learning Representations (ICLR), 2020
  
• Wasserstein adversarial examples
  Eric Wong, Frank R. Schmidt, J. Zico Kolter
  In Proceedings of the International Conference on Machine Learning (ICML), 2019
  source code on Github
• Scaling provable adversarial defenses
  Eric Wong, Frank R. Schmidt, Jan Hendrik Metzen, J. Zico Kolter
  In Neural Information Processing Systems (NeurIPS), 2018
  source code on Github
• Provable defenses against adversarial examples via the convex outer adversarial polytope
  Eric Wong, J. Zico Kolter
  In Proceedings of the International Conference on Machine Learning (ICML), 2018
  Best defense paper at NIPS 2017 ML & Security Workshop
  source code on Github Blog post
• A Semismooth Newton Method for Fast, Generic Convex Programming
  Alnur Ali*, Eric Wong*, J. Zico Kolter
  In Proceedings of the International Conference on Machine Learning (ICML), 2017
  source code on Github
• An SVD and Derivative Kernel Approach to Learning from Geometric Data
  Eric Wong, J. Zico Kolter
  In Proceedings of the Conference on Artificial Intelligence (AAAI), 2015

Teaching Experience

Teaching Assistant

• Carnegie Mellon University
  • 15-388/688 Practical Data Science, Fall 2016
  • 10-715 Advanced Introduction to Machine Learning, Fall 2015
  • 15-451 Algorithms, Fall 2014
  • 08-766/08-781 Pervasive and Mobile Computing Services, Spring 2014
  • 08-766/08-781 Pervasive and Mobile Computing Services, Spring 2013
  • 08-723 Mobile Development for iOS and Android, Spring 2013

Tensor Differential Calculus

In my quest to simplify the current state of matrix differential calculus, I have been working on a more elegant, simplified form of taking derivatives of functions regardless of the input and output shape, with more direct implementation. This has resulted in the following series, which continues to grow as I use what I've learned so far.

1. Tensors
2. Differentials and their Tensor generalization
3. Simplified Backprop
4. Backprop through Matrix Inverse

Miscellaneous

I am an avid dancer in the ballroom, latin, and hip hop styles.

I am also an enthusiastic cook, and love experimenting with different ingredients. I occasionally post pictures of my creations on Instagram