My research sits at the intersection of machine learning, natural language processing, and computational social science. My thesis is focused on how AI models represent and coordinate multiple value systems, or world models. I approach this question from a both mechanistic and statistical (coming soon) lens. My curiosity also extends to exploring questions in computational social science. More about me here
Hi, I'm Aviral Chawla 👋
I study statistical methods on natural language, and how language models organize diverse world views.
Ph.D. Candidate in Complex Systems & Data Science at the Vermont Complex Systems Institute, University of Vermont. Co-advised by Juniper Lovato in Computational Ethics Lab and Sam Zhang in Science & Humanity Lab.
Research Interests
Projects
Publications
A. Chawla, G. Hall, J. Lovato. "MetaOthello: A Controlled Study of Multiple World Models in Transformers." arXiv:2602.23164 (February 2026).
B. Antonczak, M. Fay, A. Chawla, G. Rowangould. "Comprehensive and spatially detailed passenger vehicle and truck traffic volume data for the United States estimated by machine learning." Data in Brief 64 (2026) 112451.
A. Chawla, N. Cheney. "Neighbor-Hop Mutation for Genetic Algorithm in Influence Maximization." GECCO '23 Companion, Lisbon, Portugal, pp. 187–190 (July 2023).