Dec 12

Computer Science Seminar: Gaurav Mahajan (Yale University)

-
Milstein 912 & Zoom
  • Add to Calendar 2025-12-12 11:00:00 2025-12-12 12:00:00 Computer Science Seminar: Gaurav Mahajan (Yale University) Speaker: Gaurav Mahajan (Yale University) Title: Learning Low-Rank Language Models Using Conditional Queries The seminar will be available for in-person and Zoom participation. To participate online, please email inquiry-cs@barnard.edu to receive the Zoom link. Modern AI systems, spanning chess and Go engines to large language models, protein folding predictors, and self-driving cars, have achieved remarkable success, but at a steep price of ever-increasing demands on data and computation. This trajectory is environmentally unsustainable and raises a foundational question for computer scientists: Is this dependence on massive data and compute inherent, or can we design learning systems that are both efficient and powerful? In this talk, I will explore this question from the perspective of learning theory. Gaurav Mahajan is a Postdoctoral Researcher at the Institute for Foundations of Data Science at Yale University, mentored by Dan Spielman. His research is broadly in learning theory, and its applications to reinforcement learning, quantum computing, and evolutionary biology. He completed his Ph.D. in the theory group at UC San Diego, advised by Sanjoy Dasgupta and Shachar Lovett. He also spent some fun summers at Microsoft Research, Institute for Advanced Study and Simons Institute. Milstein 912 & Zoom Barnard College barnard-admin@digitalpulp.com America/New_York public

Speaker: Gaurav Mahajan (Yale University)

Title: Learning Low-Rank Language Models Using Conditional Queries

The seminar will be available for in-person and Zoom participation. To participate online, please email inquiry-cs@barnard.edu to receive the Zoom link.

Modern AI systems, spanning chess and Go engines to large language models, protein folding predictors, and self-driving cars, have achieved remarkable success, but at a steep price of ever-increasing demands on data and computation. This trajectory is environmentally unsustainable and raises a foundational question for computer scientists: Is this dependence on massive data and compute inherent, or can we design learning systems that are both efficient and powerful? In this talk, I will explore this question from the perspective of learning theory.


Gaurav Mahajan is a Postdoctoral Researcher at the Institute for Foundations of Data Science at Yale University, mentored by Dan Spielman. His research is broadly in learning theory, and its applications to reinforcement learning, quantum computing, and evolutionary biology. He completed his Ph.D. in the theory group at UC San Diego, advised by Sanjoy Dasgupta and Shachar Lovett. He also spent some fun summers at Microsoft Research, Institute for Advanced Study and Simons Institute.