Machine Learning Researcher
preetum@nakkiran.org
Notice: After several wonderful years at Apple, I have left to work on something new... (stay tuned!)
Past research interests: I am broadly interested in understanding deep learning. When do models generalize, and what should generalization mean? When models fail, do they fail in predictable ways? I like to identify an interesting empirical behavior (e.g. length-generalization, or calibration, or...), and then characterize the behavior as precisely as possible.
Area Chair: ICLR 2024, NeurIPS 2024, ICML 2025, ICML 2026
Calibration is one way of formalizing how models which are not Bayes-optimal can still be "good" in other ways. What types of calibration hold for deep networks, and why? Is there a meaningful and achievable notion of calibration for LLMs?
I've written a tutorial on diffusion models, studied the mechanisms of diffusion generalization (classifier-free guidance, composition), and helped develop new methods (TarFlow).
Why do overparameterized models generalize? Why do underparameterized models generalize? Are these questions related? Why do LLMs generalize out-of-distribution, and how do we formalize this?
I completed my PhD at Harvard, advised by Madhu Sudan and Boaz Barak. In my postdoc I worked with Misha Belkin. I did my undergrad in EECS at UC Berkeley. Go Bears!
I'm broadly interested in theory and science. I worked at Apple's Machine Learning Research group from 2022-2026, under Josh Susskind and Samy Bengio. In the past, I've interned at OpenAI (with Ilya Sutskever), Google Brain (with Behnam Neyshabur, Hanie Sedghi), and have done research in error-correcting codes, distributed storage, and cryptography. I'm grateful for past support from NSF GRFP, the Google PhD Fellowship, and the Simons Institute.