Gaurab Pokharel Headshot

👋 Gaurab, a PhD Student at GMU

About Me:

My hometown is Kathmandu, Nepal. I graduated Fall 2021 from Oberlin College with a degree in Computer Science (High Honors) and Mathematics. I am currently a Second Year PhD student of Computer Science at George Mason University (GMU) under the advisement of Dr. Sanmay Das.

As a researcher in AI and machine learning, my interests span across game theory and its applications in the allocation of scarce societal resources. I am particularly passionate about exploring how AI can address issues of distributive justice, and I am committed to investigating the implications of various policies and mechanisms in this context

  • Full Name : Gaurab Pokharel
  • Affiliation : George Mason University
  • Email :

Ben's Resume HTML Template

When I'm not immersed in code, you'll likely find me outside or in the kitchen experimenting with new recipes. I'm also an avid drummer and enjoy exploring different cuisines. And when I'm feeling competitive, I love to play video games - if you're into Valorant, feel free to hit me up, I'm always on the lookout for new gaming buddies!.

Ongoing Projects

Converging to Stability in Two-Sided Bandits: The Case of Unknown Preferences on Both Sides of a Matching Market

We research the problem of repeated two-sided matching without explicit communication between agents, called two-sided bandits. Recent studies created algorithms that lead to stable matchings when one side must learn their preferences, and the preferences of the other side are common knowledge, and the matching mechanism uses simultaneous proposals. Our new algorithms converge to stable matchings for two more challenging settings: one where the arm preferences are not common knowledge, and a second where the arms are uncertain about their preferences. Our algorithms use optimistic beliefs about arms' preferences and update them over time, combining beliefs with the value of matching when proposing.(Working Draft, ArXiv)

Equity, and Incentives in Resource Allocation for Homelessness Alleviation

We aim to explore how AI and algorithmic techniques can best benefit society in resource allocation, prioritization, and prediction, specifically regarding homelessness prevention. Our research will address foundational questions such as (1) using outcome predictions to allocate services to different individuals and households, (2) managing competing priorities within and between institutions and agents, and (3) evaluating the fairness of prioritization mechanisms. Our focus is on resource allocation for those experiencing or at risk of homelessness. This project is in its initial phases and shows promising results and novel insights into the issue.

Say hello!