sihuid (at) princeton (dot) edu

I am a PhD student in Electrical and Computer Engineering advised by Prateek Mittal. My research interests lie at the intersection of machine learning and security. I am broadly interested in trustworthy ML. Currently, the focus of my research is on adversarial robustness. I was also fortunate to work with IBM Research's Trustworthy AI group last summer under Payel Das, mentored by Aurelie Lozano and Subhajit Chaudhury.

Before joining Princeton, I completed my undergrad at Caltech in CS with a minor in Information and Data Sciences. During that time, I was fortunate to work with Anima Anandkumar and Yisong Yue, where I worked on projects in neuro-inspired ML and Bayesian optimization.

Publications

Preprint

Towards Multi-robust Models via Regularized Continual Robust Training

Sihui Dai*, Christian Cianfarani*, Arjun Nitin Bhagoji, Vikash Sehwag, Prateek Mittal

Position: Towards Resilience Against Adversarial Examples

Sihui Dai, Chong Xiang, Tong Wu, Prateek Mittal

Conference

Larimar: Large Language Models with Episodic Memory Control (ICML 2024)

Payel Das, Subhajit Chaudhury, Elliot Nelson, Igor Melnyk, Sarath Swaminathan, Sihui Dai, Aurélie Lozano, Georgios Kollias, Vijil Chenthamarakshan, Jiří, Navrátil, Soham Dan, Pin-Yu Chen

PatchCURE: Improving Certifiable Robustness, Model Utility, and Computation Efficiency of Adversarial Patch Defenses (USENIX 2024)

Chong Xiang, Tong Wu, Sihui Dai, Jonathan Petit, Suman Jana, Prateek Mittal

Characterizing the Optimal 0-1 Loss for Multi-class Classification with a Test-time Attacker (NeurIPS 2023 Spotlight)

Sihui Dai*, Wenxin Ding*, Arjun Nitin Bhagoji, Daniel Cullina, Ben Y. Zhao, Haitao Zheng, Prateek Mittal

MultiRobustBench: Benchmarking Robustness Against Multiple Attacks (ICML 2023)

Sihui Dai, Saeed Mahloujifar, Chong Xiang, Vikash Sehwag, Pin-Yu Chen, Prateek Mittal

Formulating Robustness Against Unforeseen Attacks (NeurIPS 2022)

Sihui Dai, Saeed Mahloujifar, Prateek Mittal

Robust Learning Meets Generative Models: Can Proxy Distributions Improve Adversarial Robustness? (ICLR 2022)

Vikash Sehwag, Saeed Mahloujifar, Tinashe Handina, Sihui Dai, Chong Xiang, Mung Chiang, Prateek Mittal

Neural Networks with Recurrent Generative Feedback (NeurIPS 2020)

Yujia Huang, James Gornet, Sihui Dai, Zhiding Yu, Tan Nguyen, Doris Tsao, Anima Anandkumar

Workshop

Parameterizing Activation Functions for Adversarial Robustness (DLS 2022)

Sihui Dai, Saeed Mahloujifar, Prateek Mittal

Robustness from Perception (ICLR Security and Safety in Machine Learning Systems Workshop 2021)

Saeed Mahloujifar, Chong Xiang, Vikash Sehwag, Sihui Dai, Prateek Mittal

Multi-task Bayesian Optimization via Gaussian Process Upper Confidence Bound (ICML Real World Experiment Design and Active Learning Workshop 2020)

Sihui Dai, Jialin Song, Yisong Yue

Teaching

COS/ECE 432 (Princeton, 2022): Information Security
CMS165 (Caltech, 2020): Foundations of Machine Learning and Statistical Inference
CS156a (Caltech, 2019): Learning Systems
CS1 (Caltech, 2017,2018): Introduction to Computer Programming
CS11 (Caltech, 2018): Computer Language Lab (C++ track)

Awards and Honors

NSF Graduate Research Fellowship (2020)
Bhansali Family Prize in Computer Science, Caltech (2020)