Sarina Xi

I'm a second year Master's student in the Machine Learning Department at Carnegie Mellon University, supervised by Nihar Shah. Broadly, I am interested in using ML to better assist and augment scientific evaluation and discovery. I completed my undergraduate degree in Engineering Science at the University of Toronto, where I was supervised by David Lie for my thesis on utility-preserving private NLP inference.

Email  /  CV  /  Scholar  /  Github

FLAWS: A Benchmark for Error Identification and Localization in Scientific Papers
Sarina Xi, Vishisht Rao, Justin Payan, Nihar B. Shah
Preprint, 2025

An automated benchmark and evaluation framework to assess the ability of LLMs to identify and localize errors in scientific papers.

Who is a Better Matchmaker? Human vs. Algorithmic Judge Assignment in a High-Stakes Startup Competition
Sarina Xi, Orelia Pi, Miaomiao Zhang, Becca Xiong, Jacqueline Ng Lane, Nihar B. Shah
IAAI, 2026   (Oral Presentation)

Developed and deployed a judge assignment similarity algorithm at the 2025 Harvard President's Innovation Challenge and showed that its performance is on par with human expert assignments.

A Gold Standard Dataset For the Reviewer Assignment Problem
Ivan Stelmakh, John Wieting, Sarina Xi, Graham Neubig, Nihar B. Shah
TMLR, 2025

Gold standard dataset to evaluate and compare similarity algorithms used for review assignment in flagship ML/CS conferences.