ANDREW ILYAS

I am a first-year PhD student at MIT, fortunate to be coadvised by Costis Daskalakis and Aleksander Madry. I recently graduated from MIT with a concurrent B.S./M.Eng, double majoring (for B.S.) in Computer Science and Math.

My master’s thesis was supervised by Costis Daskalakis at MIT CSAIL. I’m also a founding member of labsix.

In past years, I’ve worked in the MIT CBMM, the Database Lab at MIT, and as an intern at Two Sigma Labs, where I worked on Convex Optimization and Online Learning.

Publications

* denotes equal contribution

Prior Convictions: Black-box Adversarial Attacks with Bandits and Priors
Andrew Ilyas*, Logan Engstrom*, Aleksander Madry (2017)
ICLR 2019. Github

How Does Batch Normalization Help Optimization?
Shibani Santurkar*, Dimitris Tsipras*, Andrew Ilyas*, Aleksander Madry (2018)
Blog Post, Video (3 minutes)
Oral presentation, NeurIPS 2018.

Black-box Adversarial Attacks with Limited Queries and Information
Andrew Ilyas*, Logan Engstrom*, Anish Athalye*, Jessy Lin* (2017)
ICML 2018. Partial-Information/GCV Blog Post, Label-Only Attack Blog Post, Github

Synthesizing Robust Adversarial Examples
Anish Athalye*, Logan Engstrom*, Andrew Ilyas*, Kevin Kwok (2017)
ICML 2018. Blog Post

Training GANs with Optimism
Constantinos Daskalakis*, Andrew Ilyas*, Vasilis Syrgkanis*, Haoyang Zeng* (2017)
ICLR 2018. Github

Extracting Syntactic Patterns From Databases
Andrew Ilyas, Joana M.F. da Trindade, Raul C. Fernandez, Samuel Madden (2017)
ICDE 2018. Github

MicroFilters: Harnessing Twitter for Disaster Managment
Andrew Ilyas (2014)
Chairman’s award winner, IEEE GHTC 2015.

Short Papers

Evaluating and Understanding the Robustness of Adversarial Logit Pairing
Logan Engstrom*, Andrew Ilyas*, Anish Athalye* (2018)
NeurIPS Security in Machine Learning Workshop 2018

Preprints

Are Deep Policy Gradient Algorithms Truly Policy Gradient Algorithms?
Andrew Ilyas*, Logan Engstrom*, Shibani Santurkar, Dimitris Tsipras, Firdaus Janoos, Larry Rudolph (2018)
Blog Post (1/3), Blog Post (2/3)

The Robust Manifold Defense: Adversarial Training using Generative Models
Andrew Ilyas, Ajil Jalal, Eirini Asteri, Constantinos Daskalakis, Alex G. Dimakis (2017)
Blog Post

Personal Projects

In this section I have included a list of previous hardware/software projects, including hackathon products, and weekend projects.

Research Archives (Pre-MIT)

In this section I have included a near-complete list of previous research topics, and links to the papers, reports, or products created as a result of that research. The projects are sorted chronologically, with every project from 2009 (Grade 5!) to present (Grade 12).

Projects