Chantal Shaib
Hello! I'm a post-doctoral researcher at MIT, working with Marzyeh Ghassemi. I recently graduated with a PhD from Northeastern University, advised by Byron Wallace. My research focuses on evaluating linguistic variance in generative models and its downstream implications for robustness, novelty, and quality assessments. In the past, I received my MSc in Applied Computing at the University of Toronto (advised by Frank Rudzicz); and before that, I received my BSc in Neuroscience at Carleton University. 🇨🇦
I am also the developer/maintainer of diversity, a Python toolkit for (efficiently) evaluating diversity in text.
Generally, I'm interested in LLM evals, modeling language and its structure, e.g.,
-
(1) Model Robustness and Safety
(2) Human Preference and Stylistic Cues in LLM Writing
(3) Multi-document Reasoning and Synthesis
Please see my Google Scholar for an up-to-date list of my publications!
