* Denotes first-authorship.
You can also find my papers on my Google Scholar profile.
Provided the first comprehensive analysis of LLM use across the peer review pipeline, with particular attention to interaction effects: not just whether LLM-assisted papers or LLM-assisted reviews are different in isolation, but whether LLM-assisted reviews evaluate LLM-assisted papers differently.
Vibhhu Sharma*, Thorsten Joachims, Sarah Dean
Under Review [Preprint]
Conducted a comprehensive comparison between risk-based targeting and biased treatment effect-based targeting as common targeting strategies used by policymakers to determine who to treat under a fixed budget in real world settings.
Vibhhu Sharma*, Bryan Wilder
Poster at ICLR 2025 [Preprint]
A novel statistical methodology to estimate heterogeneous treatment effects when generalizing from study populations to target populations with previously unobserved covariates.
Khurram Yamin, Vibhhu Sharma, Ed Kennedy, Bryan Wilder
TMLR [Preprint]
Introducing a unified framework for auditing recommender systems from a causal lens and providing a general recipe for defining auditing metrics.
Vibhhu Sharma*, Shantanu Gupta, Nil-Jana Akpinar, Zachary Lipton, Liu Leqi
Presented at the FAccTRec Workshop at RecSys 2024 [Preprint]
A novel method to counterfactually explain recommendations from visual recommender systems.
Neham Jain, Vibhhu Sharma*, and Gaurav Sinha
Published at WWW ‘24: Companion Proceedings of the ACM Web Conference 2024 [Paper]