Gabriel Missael Barco

PhD student in Physics at Université de Montréal and Mila. Probabilistic ML for strong lensing and dark matter.

Probabilistic ML for astrophysics

Gabriel Missael Barco

I am a PhD student in Physics at Université de Montréal and Mila, advised by Yashar Hezaveh and Laurence Perreault-Levasseur. I work on probabilistic machine learning for strong gravitational lensing and dark matter.

Research Publications
Diagram of strong-lensing inference variables: source, lens, dark-matter substructure, and observation
Strong-lensing dark-matter inference.

Featured Work

Figure from the ApJ paper showing biased and corrected source reconstructions under misspecified learned priors
ApJ paper

Correcting misspecified data-driven priors

Updating learned priors when simulations bias posterior reconstructions.

arXiv DOI
Figure from the blind inversion paper showing observed lensing data and joint source-lens posterior reconstructions
NeurIPS ML4PS 2025

Blind strong-lensing inversion

Joint source and lens-mass inference from noisy strong-lensing images.

arXiv Poster
Scientific software

Caustics

Open-source tools for differentiable strong-lensing simulations.

GitHub JOSS