Francisco Villaescusa-Navarro

114 posts

Francisco Villaescusa-Navarro banner
Francisco Villaescusa-Navarro

Francisco Villaescusa-Navarro

@paco_astro

Cosmologist in the morning, deep learner in the afternoon, and dreamer at night. Father all day. Quijote. CAMELS. Research Scientist @ Simons Foundation.

New York, NY, USA Katılım Ağustos 2021
437 Takip Edilen979 Takipçiler
Francisco Villaescusa-Navarro retweetledi
Markus J. Buehler
Markus J. Buehler@ProfBuehlerMIT·
Discovery is not search - it is thermodynamics! I'll explain this & much more at the @AImeetsScience Seminar on Jan 30, 2026, including how we are teaching machines to lead this process - forcing new worlds into existence when old ones can no longer remain stable. Link in the reply! Thank you @ChenhaoTan for organizing this exciting series of talks with amazing co-presenters @boknilev @yisongyue @cgeorgiaw @HannesStaerk @borisbolliet @paco_astro
Markus J. Buehler tweet media
English
0
10
68
9.6K
Francisco Villaescusa-Navarro retweetledi
Chenhao Tan
Chenhao Tan@ChenhaoTan·
Happy new year! The AI & Scientific Discovery Seminar is returning this quarter. Last quarter was incredible, from protein design to AI scientists to automated bio labs. Huge thanks to all our amazing speakers and attendees 🙌 We’re kicking off Winter Quarter with an 🔥 lineup, starting this Friday at 11am CT! 👉 @boknilev will share how interpretability methods are driving scientific discovery. Links in the thread. @yisongyue @cgeorgiaw @HannesStaerk @borisbolliet @paco_astro
Chenhao Tan tweet media
English
2
13
50
8.7K
Francisco Villaescusa-Navarro retweetledi
Aditya P. Advani
Aditya P. Advani@aditya_advani·
Just read the Denario paper from @paco_astro and @FlatironInst - a multi-agent AI scientist with a clean API. What caught my eye: they report ~10% of outputs across domains raised an intriguing question or finding. That's a trainable signal. If you can reliably identify interesting vs. not, that's possibly an RL opportunity waiting to happen. Curious where this goes. Link to paper and code in comments.
Aditya P. Advani tweet media
English
1
2
4
214
Francisco Villaescusa-Navarro retweetledi
LangChainJP
LangChainJP@LangChainJP·
The Denario project: Deep knowledge AI agents for scientific discovery - arXiv👇 arxiv.org/abs/2510.26887
English
0
1
3
848
Francisco Villaescusa-Navarro retweetledi
Teresa Huang
Teresa Huang@TeresaNHuang·
Can we advance cosmology via ML? Can we build better benchmarks for graph learning? Introducing CosmoBench: a large-scale cosmology benchmark for graph/geometric ML, with 34k clouds and 25k trees from SOTA simulations (2PB+ data, 41M+ core hours) Now accepted to @NeurIPSConf !
English
1
5
34
5.6K
Francisco Villaescusa-Navarro retweetledi
Georgios Valogiannis
Georgios Valogiannis@GeorgiosGv89·
Excited to share our latest paper, in collaboration with @paco_astro and Marco Baldi! We perform the first application of the Wavelet Scattering Transform technique to test theories of gravity, finding very promising results! Arxiv link: arxiv.org/abs/2407.18647
English
0
2
18
1.1K
Francisco Villaescusa-Navarro
We will produce the largest collection of these simulations ever created and use AI to find new and robust observables of dark matter properties on astrophysical data. Our team contains experts on galaxy formation, cosmology, particle physics, and machine learning from more than
English
1
0
5
555
Francisco Villaescusa-Navarro
The DREAMS (DaRk mattEr and Astrophysics with Machine learning and Simulations) project just started! We are combining astrophysics, particle physics, and machine learning to learn about the nature and properties of dark matter.
Francisco Villaescusa-Navarro tweet media
English
5
7
47
3K
Francisco Villaescusa-Navarro retweetledi
CAMELS project
CAMELS project@camels_project·
The newest ‘hump’ of the CAMELS project is live! We present and make publicly available 768 hydrodynamical zoomed-in simulations of massive halos varying 5 cosmological parameters and 23 astrophysical parameters within the IllustrisTNG model. Check out arxiv.org/abs/2403.10609
CAMELS project tweet media
English
1
5
25
2.1K
Francisco Villaescusa-Navarro
Do you want to work on creating the largest collection of virtual universes to date? Would you like to explore these with state-of-the-art deep-learning techniques? Do you want to use these simulations to unveil the mysteries of the Universe? Consider applying to our 4.5 months
English
1
13
41
5.6K
Francisco Villaescusa-Navarro retweetledi
Elena Massara
Elena Massara@ElenaMassara·
🔍Worried about the presence of interlopers in your galaxy catalog? 📢We developed a new method based on GNNs to infer the interloper fraction in a catalog using likelihood-free inference. 📄 Check out our new paper: arxiv.org/abs/2309.05850
English
1
1
4
1.3K