Wenwen Si | PhD Researcher

About Me

I am a fifth-year PhD student at the University of Pennsylvania. My research focuses on developing trustworthy machine learning systems through uncertainty quantification and robust optimization, primarily from a statistical, distribution-free perspective. Previously, I received my Bachelor of Engineering in Automation from the Institute of Information Science at Tsinghua University in 2018, and my Master of Science in Computer Vision from the Robotics Institute at Carnegie Mellon University in 2019.

Research Interests

Distribution-free uncertainty quantification
Robust optimization
Trustworthy machine learning for healthcare

I'm open to collaborations or talking about anything related to ML and research -- I especially enjoy mentoring younger students interested in getting involved in the field! Feel free to reach out at wenwens@seas.upenn.edu.

Selected Publications

Reliable and Interpretable Visual Field Progression Prediction with Diffusion Models and Conformal Risk Control
Wenwen Si*, Vivian Lin*, Bo Sun, Kuk Jin Jang, Rubo Xing, Almiqdad Saeed, Nagatani Rina, Oleg Sokolsky, Insup Lee, Lama Al-Aswad
The 28th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2025.
The ARVO Annual Meeting, 2025. [PDF] [Code]
Training-Set Conditionally Valid Prediction Sets with Right-Censored Data
Wenwen Si, Hongxiang Qiu
Under review, 2025. [PDF]
ICQS: An In-Context Metric for LLM Quality Scoring
Wenwen Si, Kaustubh Sridhar, Insup Lee
Under review, 2025. [PDF]
PAC Prediction Sets Under Label Shift
Wenwen Si, Sangdon Park, Insup Lee, Edgar Dobriban, Osbert Bastani
International Conference on Learning Representations (ICLR), 2024. [PDF] [Code]
Angelic Patches for Improving Third-Party Object Detector Performance
Wenwen Si, Shuo Li, Sangdon Park, Insup Lee, Osbert Bastani
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023. [PDF] [Code]

Professional Service

I have served as a student reviewer for leading international conferences, including NeurIPS, ICML, ICLR, AISTATS, KDD, AAAI, ICCPS, ICASSP, ICPR, IJCB, MIPR, as well as prominent academic journals such as TNNLS and Neural Computing and Applications, primarily in the domains of uncertainty quantification and robust computer vision.

Invention Disclosures

Pattern-Guided Diffusion Models
Invention disclosure filed, University of Pennsylvania, 2025.
Status: Under review. Details available upon request.
Angelic Patches for Safe Third-Party Object Detection
Invention disclosure filed, University of Pennsylvania, 2022.
Status: Under review. Details available upon request.