About
I am a Ph.D. candidate in Statistics at the University of Michigan, advised by Ambuj Tewari. My research focuses on building more reliable AI systems by asking how we can recognize when a model is uncertain or unreliable and use that information to make better downstream decisions.
I have explored these questions across language modeling, autonomous driving, and scientific machine learning. In 2026, I was a Data Science Research Intern at Waymo.
Current research
- Learning to Partially Defer for Sequences — selective delegation for long structured outputs
- Reasoning Models for Protein Disorder Prediction — post-training and evaluation of biological reasoning
- A Greedy PDE Router — routing between classical solvers and neural operators
- Conformal Prediction for Robust Decision-Making — uncertainty-aware optimization and control
Publications
- A Greedy PDE Router for Blending Neural Operators and Classical Methods
Sahana Rayan, Yash Patel, and Ambuj Tewari · Submitted to ICLR 2027
Link - Learning to Partially Defer for Sequences
Sahana Rayan and Ambuj Tewari · Submitted to TMLR
Link - Conformal Robust Control of Linear Systems
Yash Patel, Sahana Rayan, and Ambuj Tewari · AISTATS 2026
Link - Conformal Contextual Robust Optimization
Yash Patel, Sahana Rayan, and Ambuj Tewari · AISTATS 2024
Link
Experience
Waymo, Data Science Research Intern (Summer 2026)
Worked on evaluation and sampling methods for autonomous-driving perception analysis.University of Michigan, Doctoral Researcher (2022–present)
Research in reliable machine learning, uncertainty quantification, learning to defer, foundation models, and scientific machine learning.University of California, Berkeley, Undergraduate Research Fellow (2021–2022)
Worked on replay-based methods for continual learning with Banghua Zhu and Jiantao Jiao.
Talks
- International Joint Conference on Computational and Financial Econometrics (CFE) and Computational and Methodological Statistics (CMStatistics), 2025
Link - Michigan Student Symposium for Interdisciplinary Statistical Sciences, 2025
Link
Awards
- Best Oral Presentation, Michigan Student Symposium for Interdisciplinary Statistical Sciences, 2025
- Ph.D. Student Service Award, University of Michigan Department of Statistics, 2024
- Outstanding Graduate Student Instructor Team Award, Honorable Mention, 2023
Teaching
- Statistics and Artificial Intelligence — Winter 2023, Fall 2023, Winter 2024, Fall 2024, and Fall 2025
- Bayesian Data Analysis — Winter 2025
- Introduction to Statistics and Data Analysis — Fall 2022