People

Gaurav Mahajan joined FDS in the summer of 2023, the first year of our postdoc program. Gaurav earned his PhD in Computer Science at the University of California, San Diego in 2022. During his time as a graduate student, he worked in learning theory and reinforcement learning theory. A central theme of his work was understanding the differences between computational complexity (how much computation we need) and statistical complexity (how much data we need) of learning in sequential models like playing Atari games and generative language models. He has also spent some rewarding summers at Microsoft Research, The Institute for Advanced Study and The Simons Institute. As an FDS postdoc, Gaurav is focusing on the areas of quantum computation and discrepancy theory. In the area of quantum computation, he is currently working on understanding the role of entangled measurements in learning quantum states. In discrepancy theory, he is thinking about one-sided sparsity: the discrepancy of column sparse matrices.

Edit profile