Janet Jhih-Yi Hsieh | Carnegie Mellon University

About Me

I graduated from Carnegie Mellon University with a Master’s in Computer Science where I was fortunate to have been advised by Nihar Shah and Aditi Raghunathan. Prior to that, I graduated from Carnegie Mellon University with a Bachelor’s in Computer Science where I was fortunate to be advised by Virginia Smith and Tian Li for my senior thesis.

Research Interests

I’m interested in pursuing responsible AI designs in three ways: (1) the safety and fairness of AI models and their down-stream deployments, (2) beneficial human-AI collaboration, and (3) technical AI governance.

News

[September 2025] I will be presenting a poster of our work on vulnerabilities in peer reviewer matching at the International Congress on Peer Review and Scientific Publication, 2025.
[August 2025] I will be speaking at the USENIX Security Symposium 2025 about our accepted paper.
[June 2025] Our paper about vulnerabilities in peer reviewer matching is accepted to USENIX Security 2025.

Publications

Vulnerability of Text-Matching in ML/AI Conference Reviewer Assignments to Collusions

Jhih-Yi (Janet) Hsieh, Aditi Raghunathan, Nihar B. Shah

The USENIX Security Symposium (USENIX Sec), 2025.

(Non-archival) Championing Open-source DEvelopment in Machine Learning Workshop at ICML (CODEML), 2025.

(Long abstract) International Congress on Peer Review and Scientific Publication, 2025.

(Poster) NSF AI-SDM Workshop on Human-AI Complementarity for Decision Making, 2025.

PDF Code

Mathematical Justification of Hard Negative Mining via Isometric Approximation Theorem

Albert Xu, Jhih-Yi Hsieh, Bhaskar Vundurthy, Nithya Kemp, Eliana Cohen, Lu Li, Howie Choset

The International Conference on Learning Representations (ICLR), 2024.

PDF

On Trade-Offs Between Fairness, Robustness, and Privacy Through Tilted Losses

Jhih-Yi Hsieh, Advised by Tian Li and Virginia Smith

Undergraduate Thesis, Technical Report, 2023

PDF