I am a PhD Research Intern at Microsoft. I am part of Jaron Lanier's group, and work on analyzing and improving the performance of advanced algorithms on large-scale datasets and cutting-edge research in machine intelligence and machine learning applications. In particular our group focuses on the development of a "Theory Explorer," a machine learning algorithm to explore theory landscape manifolds. Of particular interest to us is the landscape of infinite matrix models as Beyond Standard Model theories.
I am also a third year graduate student at the University of California, Berkeley, pursuing a PhD in Physics. I am a Graduate Student Research Affiliate at Lawrence Berkeley National Lab, advised by Benjamin Nachman, who leads the Machine Learning Group in the Physics Division. I work on using Deep Learning, specifically generative models to develop methods that may be applied to high energy physics data. While the work I do is inspired by high energy physics problems, the methods I develop are data agnostic, and can be applied to diverse data sets to infer interesting properties therein. For example, while we developed our symmetry discovery algorithm SymmetryGAN because learning the symmetries of a system is valuable physics insight, SymmetryGAN could well be applied to all kinds of datasets from financial markets to social networks to glean valuable insights about the large scale structure of that dataset.
I graduated from Yale in 2020 with simultanous Bachelor of Science and Master of Science degrees with Distinction, having double majored in Mathematics (Intensive) and Physics (Intensive). When I'm not thinking about physics, you may find me playing chess or the piano.
Publications
1. Krish Desai, Benjamin Nachman, and Jesse Thaler, SymmetryGAN, Phys. Rev. D 105, 096031(2022)
2. Krish Desai, Benjamin Nachman, and Jesse Thaler, Symmetry Discovery with Deep Learning, NeurIPS ML4PS 117 (2021)
3. Adelstein, I., Desai, K., Ji, A. et al. Oblivious points on translation surfaces. J. Geom. 113, 6 (2022). https://doi.org/10.1007/s00022-021-00620-4
Machine Learning for High Energy Physics