Research

EigenBench: A Comparative Behavioral Measure of Value Alignment

Jonathn Chang, Leonhard Piff, Suvadip Sana, Jasmine X. Li, Lionel Levine

Oral · ~top 1%Fourteenth International Conference on Learning Representations (ICLR), 2026

Paper

Pluralistic Preference Alignment via Sortition-Weighted RLHF

Suvadip Sana, Jinzhou Wu, Martin T. Wells

ICML 2026, Pluralistic Alignment Workshop

Paper

Gradation of Arrow's Axioms

Suvadip Sana, Peter Rock, Martin T. Wells, Moon Duchin

Submitted to Annals of Applied Statistics

Preprint

Mixing Times of Glauber Dynamics on Masked Language Models

Suvadip Sana, Sami Wolf, Neer Mehta, Alina Shah, Aitzaz Shaikh, Janna Goodman, Lionel Levine

Submitted to NeurIPS 2026

Preprint

Statistical Aspects of Sortition

Suvadip Sana, Martin T. Wells

In preparation

Quantitative Relaxations of Arrow's Axioms

Suvadip Sana, Daniel Brous, Martin T. Wells, Moon Duchin

Preliminary version of "Gradation of Arrow's Axioms"

arXiv:2506.12961

A scalar matching factor on the Birkhoff polytope characterizing permutation and uniform matrices

Suvadip Sana

arXiv:2406.16284