Jay Agarwal

Hi! I'm a 5th year dual degree student in Physics and Computer Science at BITS Pilani. I'm interested in AI for physics, especially applications in particle physics and AMO physics. I also really enjoy studying physics and maths more broadly, particularly group theory and particle physics.

This website is perpetually a work in progress. So, please don't mind if something is not working :)

email github cv linkedin

recent

How Far Can Foundation Models Be Distilled for Efficient Jet Tagging? - Preprint, 2026
What Do Lorentz-Equivariant Jet Taggers Learn? — ICML, 2026 · AI4Physics Workshop

experience

I'm currently a visiting Researcher at the KEK Theory Center, Japan, working on AI for high-energy physics as part of my thesis at BITS Pilani under the supervision of Prof. Mihoko Nojiri.

I work with Prof. Dhruv Kumar in the PhantomGrad lab at BITS on mechanistic interpretability and distillation of symmetry-aware models for the LHC, which sent me down a rabbit hole into mech interp more broadly.

I'm also part of the TAMBO Collaboration, working with Tommaso Dorigo and the MODE Collaboration on optimizing detector layouts using differentiable programming for the TAMBO neutrino observatory.

Previously, I worked on atomic frequency standards and coherent population trapping at DRDO with Dr. Poornendu Chaturvedi, laser-assisted radiative recombination with Prof. Amol Holkundkar at BITS, and causal inference as a Quantitative Analyst at Wells Fargo.