I research ways in which the tools and perspective from theoretical physics can be applied to artificial intelligence. In particular, I co-authored of The Principles of Deep Learning Theory with Sho Yaida (also based on research in collaboration with Boris Hanin), which will be published by Cambridge University Press in early 2022. (https://deeplearningtheory.com)
I am a Research Affiliate at the Center for Theoretical Physics at MIT and also a Principal Researcher at Salesforce, having arrived via acquisition of Diffeo where I was Co-Founder and Chief Technology Officer. I am also an Affiliate of the NSF AI Institute for Artificial Intelligence and Fundamental Interactions (IAIFI).
Sho and I want to thank jylin04 for this really nice post and endorse the distillation of our key results in her 8-page summary. We also agree that it would be interesting to make further connections between our work—in particular the effective theory framework—and interpretability, and we’d be really glad to explore and discuss that further.