cat-rex

🌱I’m Elisa Tsai. I am a final year Ph.D. candidate in Computer Science and Engineering at the University of Michigan. I am fortunate to be advised by Professor Atul Prakash.

My research focuses on building pragmatic, GenAI-powered systems for web security. I also design algorithms for data efficiency and inference efficiency for vision and large language models.

Outside of work, I occasionally record music (Soundcloud, 网易云), and once every few years, I draw stuff. There’s also an archive of my Chinese blogs.

My Chinese given name is 文韬 (wén tāo), which comes from the Chinese idiom 文韬武略 (wén tāo wǔ lüè), meaning good at writing and good at fighting 😁

Links and Contact Info: [C.V.] [Google Scholar] [Github] [LinkedIn]

Selected Publications

Harmful Terms and Where to Find Them: Measuring and Modeling Unfavorable Financial Terms and Conditions in Shopping Websites at Scale
Elisa Tsai, Neal Mangaokar, Boyuan Zheng, Haizhong Zheng, Atul Prakash
To appear, The Web Conference (WWW) 2025 (Oral)
ELFS: Label-Free Coreset Selection with Proxy Training Dynamics
Haizhong Zheng (co-lead), Elisa Tsai (co-lead), Yifu Lu, Jiachen Sun, Brian R. Bartoldson, Bhavya Kailkhura, Atul Prakash
To appear, The International Conference on Learning Representations (ICLR) 2025
Terms of Deception: Exposing Obscured Financial Obligations in Online Agreements with Deep Learning
Elisa Tsai, Anoop Singhal, Atul Prakash
Deep Learning Security and Privacy Workshop (DLSP) 2024
Detecting Social Engineering Scams While Preserving User Privacy in the Digital Era (Proposal Position Paper)
Atul Prakash, Shivani Kumar, Elisa Tsai
Workshop on Technology and Consumer Protection (ConPro) 2024

For the full publication list, please check out my Google Scholar.

Teaching

Winter 2024 Computer & Network Security Grad Student Instructor. This is a seminar-based course with an extensive reading list on various computer and network security topics.

Fall 2023 EECS 598-012 Secure and Trustworthy ML Grad Student Instructor. This class discusses different types of adversarial machine learning attacks and their corresponding defenses, as well as model stealing, poisoning, fairness, and privacy.

Fall 2023 EECS 281 Data Structures and Algorithms Grad Student Instructor.

Community Service

University of Michigan

  1. SECRIT (SECurity Reading Is Terrific) Reading Group Co-Host
    I hosted SECRIT from 2021-2024, where we discuss recent papers on security and privacy. [Wiki] [Calendar]

  2. CSEG Outreach Chair
    I was the Outreach Chair for CSEG (CSE Grad) for year 2022-2023. We organize outreach events and workshops such as Xplore Engineering and Discover Engineering

  3. CSEG Social Co-Chair
    I was one of the co-chairs for CSEG social events in 2022-2023.

Reviewing Service

  1. Program Committee, WWW’25 Short Paper Track
  2. Reviewer, WWW’25 Research Track
  3. Reviewer, ICLR 2025 SCOPE Workshop
  4. Reviewer, ICLR 2025


Last Updated: