Scholar
Leopoldo Sarra
Google Scholar ID: OVQZzUEAAAAJ
Axiomatic AI
Citations & Impact
Citations
113
H-index
5
i10-index
3
Publications
18
Co-authors
0
Publications
Resume (English only)
Academic Achievements
- Developed foundation models for scientific applications such as AstroCLIP and the Omnimodal Foundation Model for Astronomical Sciences; investigated a technique to find collective variables based on their information content; explored Bayesian experimental design and active learning in complex high-dimensional situations, particularly in quantum many-body platforms.
Research Experience
- Currently Lead AI Scientist at Axiomatic AI, working on novel AI techniques that are interpretable and verifiable, specifically designed for scientific and engineering applications, particularly leading efforts on formal methods for science using Lean4 and AI agents. Formerly, a Flatiron Research Fellow at the Flatiron Institute, in the Foundation Models for Science initiative, where he explored building AI models more generally suited for scientific applications, especially in astrophysics, including AstroCLIP and the Omnimodal Foundation Model for Astronomical Sciences.
Education
- Received a Ph.D. in Physics from Friedrich-Alexander Universität in Erlangen, Germany, in association with the Max Planck Institute for the Science of Light. Previously, earned B.Sc. and M.Sc. degrees in Physics at Sapienza University of Rome, Italy, focusing on the statistical physics of spin glasses.
Background
- Research interests lie at the intersection of physics and artificial intelligence. Focuses on using machine learning to support scientific discovery, exploring how a future 'Artificial Scientist' could automatically learn from observations, understand relevant concepts, build new physical models, and design new experiments.
Co-authors