H₀ · 2024
Hubble constant from dark sirens
Using AI-based catalogs, he obtained the first and, at the time, most constraining Hubble constant measurement from dark sirens, during the fourth observing run of gravitational-wave detectors.
Brazilian Center for Physics Research
Clécio R. Bom
Artificial intelligence applied to physics.
Pushing the frontiers of artificial intelligence in physics. From astrophysics to geophysics.
Professor and Technology Researcher at CBPF, where he leads the Laboratory for Artificial Intelligence Applied to Physics (LAB-IA). His current work includes inverse modelling with Neural Posterior Estimation, Physics-Informed Neural Networks and agentic AI for problems in astrophysics and geophysics.
Highlights
GRB 230307
Neutron star merger
Co-author of the discovery of GRB 230307, a neutron star merger candidate.
S-PLUS
S-PLUS spokesperson
Spokesperson of the Southern Photometric Local Universe Survey and PI of the S-PLUS Transient Extension Program (STEP) and its gravitational-wave branch (STEP-GW).
Vera Rubin
Leadership at the Vera Rubin Observatory
Coordinates the TVS taskforce on LLMs and foundation models and the SLSC Strong Lensing Data Challenge (2024–2025).
2012 · 2020
Pioneering strong lensing work
One of the earliest applications of neural networks to strong gravitational lensing (2012) and the winning solution of the II worldwide lens finding challenge (2020).
Recognition
2025
Jacob Palis Award, from the Fulbright Commission and the Brazilian Academy of Sciences.
2024
Researcher Prize at the Brazil Conference at Harvard and MIT.
2024
Elected Affiliate Member of the Brazilian Academy of Sciences.
Research
Artificial intelligence methods to extract science from large astronomical and geophysical datasets.
Astrophysics and cosmology
- Search for gravitational-wave counterparts
- Cosmology with gravitational waves
- Kilonova and binary black hole astrophysics
- Supernovae and fast transients
- Search and modelling of strong lensing systems with Deep Learning
- Photometric redshifts with AI density estimators
- Galaxy morphology and search for rare objects
AI for physics and geophysics
- Foundation models, LLMs and LRMs for physics
- Velocity model building from seismic images with physics-informed AI
- Artificial intelligence assessment of seismic images
Laboratory for Artificial Intelligence Applied to Physics
LAB-IA brings together research, development and innovation at CBPF, with high-performance computing infrastructure dedicated to artificial intelligence and projects ranging from astrophysics to industry.
In the press
AI system created in Brazil will help a NASA telescope discover what changes in the sky
The system will receive the telescope’s detections, analyze the data within seconds and classify the events to help researchers identify which phenomena deserve further investigation.
Read story
Join the AGN Lensing Challenge in the LSST Era
An opportunity for researchers and students to test their approaches and contribute to new solutions for analyzing gravitational lenses in the context of large astronomical surveys. Visit the website, learn about the challenge and take part.
Read story
Who are the Brazilian researchers playing a decisive role in operating the space telescope NASA launches this Sunday
A next-generation instrument, Roman has a field of view at least 100 times larger than Hubble’s and was designed to address some of the mysteries of the universe.
Read story
Materials
Lectures, notebooks, models and code, openly shared.
Teaching
Courses on Big Data, Astroinformatics, Deep Learning and Physics, with slides, problem sets and tasks.
ExploreMinicourses
Cosmology with Python and Deep Learning at the CBPF Schools, with slides, videos and notebooks.
ExploreScripts & Models
Trained models, catalogs and student projects available for reproduction.
Explore