Publications

(*) indicates equal contribution

(α-β) indicates alphabetical order

You can also find my work on my Google Scholar profile.

Preprints

How Many Repeated Pairwise Comparisons Are Needed for Ranking Under Heterogeneity?

(α-β) Shashaank Aiyer and Han Shao

arXiv 2026

Active Learning via Regression Beyond Realizability

Atul Ganju, Shashaank Aiyer*, Ved Sriraman*, and Karthik Sridharan

arXiv 2025

Conference Publications

Scale-Sensitive Shattering: Learnability and Evaluability at Optimal Scale

(α-β) Shashaank Aiyer, Yishay Mansour, Shay Moran, Han Shao, and Tom Waknine

NeurIPS 2026. Oral Presentation

A Theoretical Framework for Statistical Evaluability of Generative Models

(α-β) Shashaank Aiyer, Yishay Mansour, Shay Moran, and Han Shao

ICML 2026.