arXiv subjects

Pablo Villanueva-Domingo

Publications and source records attributed to Pablo Villanueva-Domingo.

At least 19 recordsLinked to original sources

arXiv2025

The Denario project: Deep knowledge AI agents for scientific discovery

We present Denario, an AI multi-agent system designed to serve as a scientific research assistant. Denario can perform many different tasks, such as generating ideas, checking the literature, developing research plans, writing and executing code, making plots, and drafting and reviewing a scientific paper. The system has a modular architecture, allowing it to handle specific tasks, such as generating an idea, or carrying out end-to-end scientific analysis using Cmbagent as a deep-research backend. In this work, we describe in detail Denario and its modules, and illustrate its capabilities by presenting multiple AI-generated papers generated by it in many different scientific disciplines such as astrophysics, biology, biophysics, biomedical informatics, chemistry, material science, mathematical physics, medicine, neuroscience and planetary science. Denario also excels at combining ideas from different disciplines, and we illustrate this by showing a paper that applies methods from quantum physics and machine learning to astrophysical data. We report the evaluations performed on these papers by domain experts, who provided both numerical scores and review-like feedback. We then highlight the strengths, weaknesses, and limitations of the current system. Finally, we discuss the ethical implications of AI-driven research and reflect on how such technology relates to the philosophy of science. We publicly release the code at https://github.com/AstroPilot-AI/Denario. A Denario demo can also be run directly on the web at https://huggingface.co/spaces/astropilot-ai/Denario, and the full app will be deployed on the cloud.

cs.AI↗

arXiv2025

Open Source Planning & Control System with Language Agents for Autonomous Scientific Discovery

We present a multi-agent system for automation of scientific research tasks, cmbagent (https://github.com/CMBAgents/cmbagent). The system is formed by about 30 Large Language Model (LLM) agents and implements a Planning & Control strategy to orchestrate the agentic workflow, with no human-in-the-loop at any point. Each agent specializes in a different task (performing retrieval on scientific papers and codebases, writing code, interpreting results, critiquing the output of other agents) and the system is able to execute code locally. We successfully apply cmbagent to carry out a PhD level cosmology task (the measurement of cosmological parameters using supernova data) and evaluate its performance on two benchmark sets, finding superior performance over state-of-the-art LLMs. The source code is available on GitHub, demonstration videos are also available, and the system is deployed on HuggingFace and will be available on the cloud.

cs.AI↗

arXiv2023

Robust Field-level Likelihood-free Inference with Galaxies

We train graph neural networks to perform field-level likelihood-free inference using galaxy catalogs from state-of-the-art hydrodynamic simulations of the CAMELS project. Our models are rotational, translational, and permutation invariant and do not impose any cut on scale. From galaxy catalogs that only contain $3$D positions and radial velocities of $\sim 1, 000$ galaxies in tiny $(25~h^{-1}{\rm Mpc})^3$ volumes our models can infer the value of $Ω_{\rm m}$ with approximately $12$ % precision. More importantly, by testing the models on galaxy catalogs from thousands of hydrodynamic simulations, each having a different efficiency of supernova and AGN feedback, run with five different codes and subgrid models - IllustrisTNG, SIMBA, Astrid, Magneticum, SWIFT-EAGLE -, we find that our models are robust to changes in astrophysics, subgrid physics, and subhalo/galaxy finder. Furthermore, we test our models on $1,024$ simulations that cover a vast region in parameter space - variations in $5$ cosmological and $23$ astrophysical parameters - finding that the model extrapolates really well. Our results indicate that the key to building a robust model is the use of both galaxy positions and velocities, suggesting that the network have likely learned an underlying physical relation that does not depend on galaxy formation and is valid on scales larger than $\sim10~h^{-1}{\rm kpc}$.

astro-ph.CO↗

arXiv2023

Boosting the 21 cm forest signals by the clumpy substructures

We study the contribution of subhalos to the 21 cm forest signal. The halos can host the substructures and including the effects of those small scale clumps can potentially boost the 21 cm optical depth in favor of detecting the 21 cm forest signals. We estimate the boost factor representing the ratio of the optical depth due to the subhalo contribution and that due to the host halo alone (without subhalos). Even though the optical depth boost factor is negligible for a small host halo with the mass of order $10^5 M_{\odot}$, the subhalo contribution can enhance the optical depth by an order of magnitude for a host halo of order $10^7 M_{\odot}$. The resultant 21 cm absorption line abundance which is obtained by integrating over the halo mass range relevant for the 21 cm forest signal can be enhanced by up to of order $10\%$ due to the substructures. The larger boost factor for a larger host halo would be of particular interest for the 21 cm forest detection because the the contribution of the larger host halos to the 21 cm forest signals is smaller due to their higher temperature and less abundance than the smaller host halos. The subhalos hence can well help the larger host halos more important for the signal estimation which, without considering the subhalos, may not give appreciable contribution to 21 cm forest signals.

astro-ph.CO↗

arXiv2023

Inferring halo masses with Graph Neural Networks

Understanding the halo-galaxy connection is fundamental in order to improve our knowledge on the nature and properties of dark matter. In this work we build a model that infers the mass of a halo given the positions, velocities, stellar masses, and radii of the galaxies it hosts. In order to capture information from correlations among galaxy properties and their phase-space, we use Graph Neural Networks (GNNs), that are designed to work with irregular and sparse data. We train our models on galaxies from more than 2,000 state-of-the-art simulations from the Cosmology and Astrophysics with MachinE Learning Simulations (CAMELS) project. Our model, that accounts for cosmological and astrophysical uncertainties, is able to constrain the masses of the halos with a $\sim$0.2 dex accuracy. Furthermore, a GNN trained on a suite of simulations is able to preserve part of its accuracy when tested on simulations run with a different code that utilizes a distinct subgrid physics model, showing the robustness of our method. The PyTorch Geometric implementation of the GNN is publicly available on Github at https://github.com/PabloVD/HaloGraphNet

astro-ph.CO↗

arXiv2023

Learning cosmology and clustering with cosmic graphs

We train deep learning models on thousands of galaxy catalogues from the state-of-the-art hydrodynamic simulations of the CAMELS project to perform regression and inference. We employ Graph Neural Networks (GNNs), architectures designed to work with irregular and sparse data, like the distribution of galaxies in the Universe. We first show that GNNs can learn to compute the power spectrum of galaxy catalogues with a few percent accuracy. We then train GNNs to perform likelihood-free inference at the galaxy-field level. Our models are able to infer the value of $Ω_{\rm m}$ with a $\sim12\%-13\%$ accuracy just from the positions of $\sim1000$ galaxies in a volume of $(25~h^{-1}{\rm Mpc})^3$ at $z=0$ while accounting for astrophysical uncertainties as modelled in CAMELS. Incorporating information from galaxy properties, such as stellar mass, stellar metallicity, and stellar radius, increases the accuracy to $4\%-8\%$. Our models are built to be translational and rotational invariant, and they can extract information from any scale larger than the minimum distance between two galaxies. However, our models are not completely robust: testing on simulations run with a different subgrid physics than the ones used for training does not yield as accurate results.

astro-ph.CO↗

arXiv2022

Current and future neutrino limits on the abundance of primordial black holes

Primordial black holes (PBHs) formed in the early Universe are sources of neutrinos emitted via Hawking radiation. Such astrophysical neutrinos could be detected at Earth and constraints on the abundance of comet-mass PBHs could be derived from the null observation of this neutrino flux. Here, we consider non-rotating PBHs and improve constraints using Super-Kamiokande neutrino data, as well as we perform forecasts for next-generation neutrino (Hyper-Kamiokande, JUNO, DUNE) and dark matter (DARWIN, ARGO) detectors, which we compare. For PBHs less massive than $\sim \textrm{few} \times 10^{14}$ g, PBHs would have already evaporated by now, whereas more massive PBHs would still be present and would constitute a fraction of the dark matter of the Universe. We consider monochromatic and extended (log-normal) mass distributions, and a PBH mass range spanning from $10^{12}$ g to $\sim 10^{16}$ g. Finally, we also compare our results with previous ones in the literature.

hep-ph↗

arXiv2022

Robust field-level inference with dark matter halos

We train graph neural networks on halo catalogues from Gadget N-body simulations to perform field-level likelihood-free inference of cosmological parameters. The catalogues contain $\lesssim$5,000 halos with masses $\gtrsim 10^{10}~h^{-1}M_\odot$ in a periodic volume of $(25~h^{-1}{\rm Mpc})^3$; every halo in the catalogue is characterized by several properties such as position, mass, velocity, concentration, and maximum circular velocity. Our models, built to be permutationally, translationally, and rotationally invariant, do not impose a minimum scale on which to extract information and are able to infer the values of $Ω_{\rm m}$ and $σ_8$ with a mean relative error of $\sim6\%$, when using positions plus velocities and positions plus masses, respectively. More importantly, we find that our models are very robust: they can infer the value of $Ω_{\rm m}$ and $σ_8$ when tested using halo catalogues from thousands of N-body simulations run with five different N-body codes: Abacus, CUBEP$^3$M, Enzo, PKDGrav3, and Ramses. Surprisingly, the model trained to infer $Ω_{\rm m}$ also works when tested on thousands of state-of-the-art CAMELS hydrodynamic simulations run with four different codes and subgrid physics implementations. Using halo properties such as concentration and maximum circular velocity allow our models to extract more information, at the expense of breaking the robustness of the models. This may happen because the different N-body codes are not converged on the relevant scales corresponding to these parameters.

astro-ph.CO↗

arXiv2022

The CAMELS project: public data release

The Cosmology and Astrophysics with MachinE Learning Simulations (CAMELS) project was developed to combine cosmology with astrophysics through thousands of cosmological hydrodynamic simulations and machine learning. CAMELS contains 4,233 cosmological simulations, 2,049 N-body and 2,184 state-of-the-art hydrodynamic simulations that sample a vast volume in parameter space. In this paper we present the CAMELS public data release, describing the characteristics of the CAMELS simulations and a variety of data products generated from them, including halo, subhalo, galaxy, and void catalogues, power spectra, bispectra, Lyman-$α$ spectra, probability distribution functions, halo radial profiles, and X-rays photon lists. We also release over one thousand catalogues that contain billions of galaxies from CAMELS-SAM: a large collection of N-body simulations that have been combined with the Santa Cruz Semi-Analytic Model. We release all the data, comprising more than 350 terabytes and containing 143,922 snapshots, millions of halos, galaxies and summary statistics. We provide further technical details on how to access, download, read, and process the data at \url{https://camels.readthedocs.io}.

astro-ph.CO↗

arXiv2021

Shedding light on Dark Matter through 21 cm Cosmology and Reionization constraints

During the last decades, our understanding of the universe has reached a remarkable level, being able to test cosmological predictions with an astonishing precision. Nonetheless, the nature, composition, mass and interactions of the Dark Matter still remain unknown, presenting one of the most intriguing conundrums in current cosmology. In this doctoral thesis, signatures of Dark Matter candidates which can leave an impact on the process of formation of structures and on the evolution of the Intergalactic Medium are studied. This thesis is organized in three parts. Part I is devoted to a broad introduction to the fundamentals, describing the state of the art of the topics considered. The basics of the $Λ$CDM are presented in Chapter 1. Chapter 2 overviews the historical progress of evidences of Dark Matter, followed by a discussion of the status and small-scale issues of the Cold Dark Matter paradigm, examining two alternative non-standard scenarios: Warm Dark Matter and Interacting Dark Matter. Chapter 3 considers Primordial Black Holes as another Dark Matter candidate, discussing the effects of accretion of surrounding matter and the enhancement of small-scale fluctuations due to the Poisson shot noise, both of which could leave an observational impact in the Intergalactic Medium. The fundamentals of the 21 cm cosmological signal are reviewed in Chapter 4, summarizing the main processes which drive the brightness temperature, and discussing its spatial fluctuations via the power spectrum. Finally, Chapter 5 is dedicated to the ionization and thermal evolution of the Intergalactic Medium during the Cosmic Dawn and the Reionization epochs. Part II includes seven original scientific articles published during the development of the PhD, which constitute the main work of this thesis. Finally, Part III contains a summary of the main results in Spanish.

astro-ph.CO↗

arXiv2021

Weighing the Milky Way and Andromeda with Artificial Intelligence

We present new constraints on the masses of the halos hosting the Milky Way and Andromeda galaxies derived using graph neural networks. Our models, trained on thousands of state-of-the-art hydrodynamic simulations of the CAMELS project, only make use of the positions, velocities and stellar masses of the galaxies belonging to the halos, and are able to perform likelihood-free inference on halo masses while accounting for both cosmological and astrophysical uncertainties. Our constraints are in agreement with estimates from other traditional methods.

astro-ph.GA↗

arXiv2021

The CAMELS Multifield Dataset: Learning the Universe's Fundamental Parameters with Artificial Intelligence

We present the Cosmology and Astrophysics with MachinE Learning Simulations (CAMELS) Multifield Dataset, CMD, a collection of hundreds of thousands of 2D maps and 3D grids containing many different properties of cosmic gas, dark matter, and stars from 2,000 distinct simulated universes at several cosmic times. The 2D maps and 3D grids represent cosmic regions that span $\sim$100 million light years and have been generated from thousands of state-of-the-art hydrodynamic and gravity-only N-body simulations from the CAMELS project. Designed to train machine learning models, CMD is the largest dataset of its kind containing more than 70 Terabytes of data. In this paper we describe CMD in detail and outline a few of its applications. We focus our attention on one such task, parameter inference, formulating the problems we face as a challenge to the community. We release all data and provide further technical details at https://camels-multifield-dataset.readthedocs.io.

cs.LG↗

arXiv2021

A brief review on primordial black holes as dark matter

Primordial black holes (PBHs) represent a natural candidate for one of the components of the dark matter (DM) in the Universe. In this review, we shall discuss the basics of their formation, abundance and signatures. Some of their characteristic signals are examined, such as the emission of particles due to Hawking evaporation and the accretion of the surrounding matter, effects which could leave an impact in the evolution of the Universe and the formation of structures. The most relevant probes capable of constraining their masses and population are discussed.

astro-ph.CO↗

arXiv2021

21 cm Forest Constraints on Primordial Black Holes

Primordial black holes (PBHs) as part of the Dark Matter (DM) would modify the evolution of large-scale structures and the thermal history of the universe. Future 21 cm forest observations, sensitive to small scales and the thermal state of the Inter Galactic Medium (IGM), could probe the existence of such PBHs. In this article, we show that the shot noise isocurvature mode on small scales induced by the presence of PBHs can enhance the amount of low mass halos, or minihalos, and thus, the number of 21 cm absorption lines. However, if the mass of PBHs is as large as $M_{\rm PBH}\gtrsim 10 \, M_\odot$, with an abundant enough fraction of PBHs as DM, $f_{\rm PBH}$, the IGM heating due to accretion onto the PBHs counteracts the enhancement due to the isocurvature mode, reducing the number of absorption lines instead. The concurrence of both effects imprints distinctive signatures in the number of absorbers, allowing to bound the abundance of PBHs. We compute the prospects for constraining PBHs with future 21 cm forest observations, finding achievable competitive upper limits on the abundance as low as $f_{\rm PBH} \sim 10^{-3}$ at $M_{\rm PBH}= 100 \, M_\odot$, or even lower at larger masses, in unexplored regions of the parameter space by current probes. The impact of astrophysical X-ray sources on the IGM temperature is also studied, which could potentially weaken the bounds.

astro-ph.CO↗

arXiv2021

Maximum amplitude of the high-redshift 21-cm absorption feature

We examine the maximum possible strength of the global 21-cm absorption dip on the Cosmic Background Radiation at high-redshift caused by the atomic intergalactic medium, when the Lyman-$α$ coupling is maximum, assuming no exotic cooling mechanisms from interactions with dark matter. This maximum absorption is limited by three inevitable factors that need to be accounted for: $(a)$ heating by energy transferred from the Cosmic Background Radiation to the hydrogen atoms via 21-cm transitions, dubbed as 21-cm heating; $(b)$ Ly$α$ heating by scatterings of Ly$α$ photons from the first stars; $(c)$ the impact of the expected density fluctuations in the intergalactic gas in standard Cold Dark Matter theory, which reduces the mean 21-cm absorption signal. Inclusion of this third novel effect reduces the maximum global 21-cm absorption by $\sim 10\%$. Overall, the three effects studied here reduce the 21-cm global absorption by $\sim 20\%$ at $z \simeq 17$.

astro-ph.CO↗

arXiv2021

Removing Astrophysics in 21 cm maps with Neural Networks

Measuring temperature fluctuations in the 21 cm signal from the Epoch of Reionization and the Cosmic Dawn is one of the most promising ways to study the Universe at high redshifts. Unfortunately, the 21 cm signal is affected by both cosmology and astrophysics processes in a non-trivial manner. We run a suite of 1,000 numerical simulations with different values of the main astrophysical parameters. From these simulations we produce tens of thousands of 21 cm maps at redshifts $10\leq z\leq 20$. We train a convolutional neural network to remove the effects of astrophysics from the 21 cm maps, and output maps of the underlying matter field. We show that our model is able to generate 2D matter fields that not only resemble the true ones visually, but whose statistical properties agree with the true ones within a few percent down to pretty small scales. We demonstrate that our neural network retains astrophysical information, that can be used to constrain the value of the astrophysical parameters. Finally, we use saliency maps to try to understand which features of the 21 cm maps is the network using in order to determine the value of the astrophysical parameters.

astro-ph.CO↗

arXiv2020

Variations in fundamental constants at the cosmic dawn

The observation of space-time variations in fundamental constants would provide strong evidence for the existence of new light degrees of freedom in the theory of Nature. Robustly constraining such scenarios requires exploiting observations that span different scales and probe the state of the Universe at different epochs. In the context of cosmology, both the cosmic microwave background and the Lyman-$α$ forest have proven to be powerful tools capable of constraining variations in electromagnetism, however at the moment there do not exist cosmological probes capable of bridging the gap between recombination and reionization. In the near future, radio telescopes will attempt to measure the 21cm transition of neutral hydrogen during the epochs of reionization and the cosmic dawn (and potentially the tail end of the dark ages); being inherently sensitive to electromagnetic phenomena, these experiments will offer a unique perspective on space-time variations of the fine-structure constant and the electron mass. We show here that large variations in these fundamental constants would produce features on the 21cm power spectrum that may be distinguishable from astrophysical uncertainties. Furthermore, we forecast the sensitivity for the Square Kilometer Array, and show that the 21cm power spectrum may be able to constrain variations at the level of ${\cal O}(10^{-3})$.

astro-ph.CO↗

arXiv2019

Constraining the primordial black hole abundance with 21cm cosmology

The discoveries of a number of binary black hole mergers by LIGO and VIRGO has reinvigorated the interest that primordial black holes (PBHs) of tens of solar masses could contribute non-negligibly to the dark matter energy density. Should even a small population of PBHs with masses $\gtrsim \mathcal{O}(M_\odot)$ exist, they could profoundly impact the properties of the intergalactic medium and provide insight into novel processes at work in the early Universe. We demonstrate here that observations of the 21cm transition in neutral hydrogen during the epochs of reionization and cosmic dawn will likely provide one of the most stringent tests of solar mass PBHs. In the context of 21cm cosmology, PBHs give rise to three distinct observable effects: ${\textit{(i)}}$ the modification to the primordial power spectrum (and thus also the halo mass function) induced by Poisson noise, ${\textit{(ii)}}$ a uniform heating and ionization of the intergalactic medium via X-rays produced during accretion, and ${\textit{(iii)}}$ a local modification to the temperature and density of the ambient medium surrounding isolated PBHs. Using a four-parameter astrophysical model, we show that experiments like SKA and HERA could potentially improve upon existing constraints derived using observations of the cosmic microwave background by more than one order of magnitude.

astro-ph.CO↗