Ari Brill

I’m an AI safety researcher working to improve humanity’s scientific knowledge of advanced AI systems. I lead the Data Models & Validation team for PIRAMID (Physics-Informed Research for Ambitious Mechanistic Interpretability Development) within Principles of Intelligence.

My research focuses on creating mathematical and empirical models to study how AI systems develop internal representations of the world. Currently, I’m investigating how data models that exhibit critical phenomena and scale-free structure can be applied to improve AI interpretability tools.

Previously, I was a NASA Postdoctoral Program Fellow at NASA Goddard Space Flight Center. At NASA, I used deep learning and statistical analysis to investigate high-energy extragalactic astrophysics, focusing on modeling the variability of gamma-ray emission powered by supermassive black holes.

I completed my PhD in Physics at Columbia in 2021. For my thesis, I investigated the extreme flux variability of very-high-energy blazars and studied deep neural networks as an analysis method for next-generation gamma-ray telescopes. Before that, I completed a B.S. in Physics at Yale in 2015.

My research is supported by Coefficient Giving and the Long-Term Future Fund (EA Funds).