arXiv subjects

Luisa Lucie-Smith

Publications and source records attributed to Luisa Lucie-Smith.

At least 19 recordsLinked to original sources

arXiv2026

Morphology of Radio Sources in Representation Space

Understanding radio source morphologies and their classification remains challenging. We previously developed a deep clustering method based on self-supervised learning to classify a subsample of radio sources from the LOFAR Two-meter Sky Survey DR2. This yielded a labelled subset used to fine-tune an ensemble of classifiers. We aim to identify rare morphological classes in a subsample of LoTSS-DR3, which contains > 13 million sources, beyond the 12 classes previously recognised in the DR2 sample. We further aim to characterise the resulting class distribution. We applied the classifier ensemble to derive class probabilities and representations for our samples. Among DR3 sources with low class probabilities, we searched for new clusters in representation space. Moreover, we use these clusters as centroids to classify the low-probability subset. We found that 88% of the sources fall into clearly separated clusters in representation space, coinciding with the DR2 clusters. Within the remaining sources, we identified representatives of six additional morphological classes, including rare morphologies like winged and FR-ambiguous sources. The number of sources in the resulting 18 classes exhibit a strongly skewed distribution, with four dominant classes constituting 69% of the DR3 sample, while rare morphologies account for only a small fraction. The identification of additional morphological classes shows the possibility of open-set recognition with representation learning in an unexplored dataset. Unlike anomaly detection, which flags individual sources as uncommon, this approach identifies representatives of novel morphological classes, providing a framework for discovering novel source populations. The highly skewed class distribution poses a fundamental challenge for constructing balanced training datasets and highlights the need for open-set approaches in future observations.

astro-ph.IM↗

arXiv2026

Interpreting "Interpretability" and Explaining "Explainability" in Machine Learning in Physics

We review the concepts of interpretability and explainability as they apply to machine learning in physics. We define interpretability as concerning the structural transparency of a model (the ability to understand or approximate its inner workings) and explainability as concerning the scientific content of a model (the ability to map it onto domain knowledge). We discuss the trade-offs each entails (interpretability vs. expressivity; explainability vs. adaptability), the contexts in which each is needed, and the intrinsic and post-hoc tools available for achieving them. Throughout, we emphasize that machine-learned models are subject to the same scientific questions as classical models, differing only in scale, and that interpretability and explainability are best understood as deliberate modeling choices rather than inherent properties. We also emphasize the importance of task specification and intervention plans as a core aspect of model design.

physics.data-an↗

arXiv2025

Unveiling the small-scale web around galaxies with miniJPAS and DESI

We present the first statistical observational study detecting filaments in the immediate surroundings of galaxies, i.e. the local web of galaxies. Simulations predict that cold gas, the fuel for star formation, is channeled through filamentary structures into galaxies. Yet, direct observational evidence for this process has been limited by the challenge of mapping the cosmic web at small scales. Using miniJPAS spectro-photometric data combined with spectroscopic DESI redshifts when available, we construct a high-density observational galaxy sample spanning 0.2 10^(10) Msun using all nearby galaxies as tracers, combined with a probabilistic adaptation of the DisPerSE algorithm designed to overcome limitations due to photometric redshift uncertainties. Our methodology is tested and validated using mock catalogues built with random forest models applied to a simulated lightcone. Besides recovering the expected increase in galaxy connectivity (defined as the number of filaments attached to a galaxy) with stellar mass, we show that our connectivity measurements agree with 3D reference estimates from the mock galaxies. Thanks to these filament reconstructions, we explore the relation between small-scale connectivity and galaxy star formation rate, finding a mild positive trend which needs to be confirmed by follow up studies with larger sample sizes. We propose galaxy connectivity to local filaments as a powerful and physically motivated metric of environment, offering new insights into the role of cosmic structure in galaxy evolution.

astro-ph.CO↗

arXiv2025

Cosmological feedback from a halo assembly perspective

The impact of feedback from galaxy formation on cosmological probes is typically quantified in terms of the suppression of the matter power spectrum in hydrodynamical compared to gravity-only simulations. In this paper, we instead study how baryonic feedback impacts halo assembly histories and thereby imprints on cosmological observables. We investigate the sensitivity of the thermal Sunyaev-Zel'dovich effect (tSZ) power spectrum, X-ray number counts, weak lensing and kinetic Sunyaev-Zel'dovich (kSZ) stacked profiles to halo populations as a function of mass and redshift. We then study the imprint of different feedback implementations in the FLAMINGO suite of cosmological simulations on the assembly histories of these halo populations, as a function of radial scale. We find that kSZ profiles target lower-mass halos ($M_{200\mathrm{m}}\sim 10^{13.1}~\mathrm{M}_\odot$) compared to all other probes considered ($M_{200\mathrm{m}}\sim 10^{15}~\mathrm{M}_\odot$). Feedback is inefficient in high-mass clusters with $\sim 10^{15} \, \mathrm{M}_\odot$ at $z=0$, but was more efficient at earlier times in the same population, with a $\sim 5$-$10\%$ effect on mass at $2 2$). These findings are tied together by noting that, regardless of redshift, feedback most efficiently redistributes baryons when halos reach a mass of $M_{\rm 200m} \simeq {10^{12.8}}\,\mathrm{M}_{\odot}$ and ceases to have any significant effect by the time $M_{\rm 200m} \simeq {10^{15}}\,\mathrm{M}_{\odot}$. We put forward strategies for minimizing sensitivity of lensing analyses to baryonic feedback, and for exploring baryonic resolutions to the unexpectedly low tSZ power in cosmic microwave background observations.

astro-ph.CO↗

arXiv2025

Classification of Radio Sources Through Self-Supervised Learning

The morphology of radio galaxies is indicative of their interaction with their surroundings, among other effects. Since modern radio surveys contain a large number of radio sources that would be impossible to analyse and classify manually, it is important to develop automatic schemes. Unlike other fields, which benefit from established theoretical frameworks and simulations, there are no such comprehensive models built for radio galaxies. This stands as a challenge to data analysis in this field and novel approaches are required. In this study, we investigate the classification of radio galaxies from the LOFAR Two-meter Sky Survey Data Release 2 (LoTSS-DR2) using self-supervised learning. Our deep clustering classification strategy involves three main steps: (i) self-supervised pre-training; (ii) fine-tuning using a labelled subsample created from the learned representations; and (iii) performing a final classification of the selected unlabelled sample. To enhance morphological information in the representations, we developed an additional random augmentation, called a random structural view (RSV). Our results demonstrate that the learned representations contain rich morphological information, enabling the creation of a labelled subsample that effectively captures the morphological diversity within the unlabelled sample. Additionally, the classification of the unlabelled sample into 12 morphological classes yields robust class probabilities. We successfully demonstrated that a subset of radio galaxies from LoTSS-DR2, encompassing diverse morphologies, can be classified using deep clustering based on self-supervised learning. The methodology developed here bridges the gap left by the absence of simulations and theoretical models, offering a framework that can readily be applied to astronomical image analyses in other bands.

astro-ph.IM↗

arXiv2025

$Λ$CDM and early dark energy in latent space: a data-driven parametrization of the CMB temperature power spectrum

Finding the best parametrization for cosmological models in the absence of first-principle theories is an open question. We propose a data-driven parametrization of cosmological models given by the disentangled 'latent' representation of a variational autoencoder (VAE) trained to compress cosmic microwave background (CMB) temperature power spectra. We consider a broad range of $Λ$CDM and beyond-$Λ$CDM cosmologies with an additional early dark energy (EDE) component. We show that these spectra can be compressed into 5 ($Λ$CDM) or 8 (EDE) independent latent parameters, as expected when using temperature power spectra alone, and which reconstruct spectra at an accuracy well within the Planck errors. These latent parameters have a physical interpretation in terms of well-known features of the CMB temperature spectrum: these include the position, height and even-odd modulation of the acoustic peaks, as well as the gravitational lensing effect. The VAE also discovers one latent parameter which entirely isolates the EDE effects from those related to $Λ$CDM parameters, thus revealing a previously unknown degree of freedom in the CMB temperature power spectrum. We further showcase how to place constraints on the latent parameters using Planck data as typically done for cosmological parameters, obtaining latent values consistent with previous $Λ$CDM and EDE cosmological constraints. Our work demonstrates the potential of a data-driven reformulation of current beyond-$Λ$CDM phenomenological models into the independent degrees of freedom to which the data observables are sensitive.

astro-ph.CO↗

arXiv2025

Cosmological Inference with Cosmic Voids and Neural Network Emulators

Cosmic Voids are a promising probe of cosmology for spectroscopic galaxy surveys due to their unique response to cosmological parameters. Their combination with other probes promises to break parameter degeneracies. Due to simplifying assumptions, analytical models for void statistics are only representative of a subset of the full void population. We present a set of neural-based emulators for void summary statistics of watershed voids, which retain more information about the full void population than simplified analytical models. We build emulators for the void size function and void density profiles traced by the halo number density using the Quijote suite of simulations for a broad range of the $Λ\mathrm{CDM}$ parameter space. The emulators replace the computation of these statistics from computationally expensive cosmological simulations. We demonstrate the cosmological constraining power of voids using our emulators, which offer orders-of-magnitude acceleration in parameter estimation, capture more cosmological information compared to analytic models, and produce more realistic posteriors compared to Fisher forecasts. We find that the parameters $Ω_m$ and $σ_8$ in this Quijote setup can be recovered to $14.4\%$ and $8.4\%$ accuracy respectively using void density profiles; including the additional information in the void size function improves the accuracy on $σ_8$ to $6.8\%$. We demonstrate the robustness of our approach to two important variables in the underlying simulations, the resolution, and the inclusion of baryons. We find that our pipeline is robust to variations in resolution, and we show that the posteriors derived from the emulated void statistics are unaffected by the inclusion of baryons with the Magneticum hydrodynamic simulations. This opens up the possibility of a baryon-independent probe of the large-scale structure.

astro-ph.CO↗

arXiv2024

Deep learning insights into non-universality in the halo mass function

The abundance of dark matter haloes is a key cosmological probe in forthcoming galaxy surveys. The theoretical understanding of the halo mass function (HMF) is limited by our incomplete knowledge of the origin of non-universality and its cosmological parameter dependence. We present a deep learning model which compresses the linear matter power spectrum into three independent factors which are necessary and sufficient to describe the $z=0$ HMF from the state-of-the-art AEMULUS emulator to sub-per cent accuracy in a $w$CDM$+N_\mathrm{eff}$ parameter space. Additional information about growth history does not improve the accuracy of HMF predictions if the matter power spectrum is already provided as input, because required aspects of the former can be inferred from the latter. The three factors carry information about the universal and non-universal aspects of the HMF, which we interrogate via the information-theoretic measure of mutual information. We find that non-universality is captured by recent growth history after matter-dark-energy equality and $N_\mathrm{eff}$ for $M\sim 10^{13} \, \mathrm{M_\odot}\, h^{-1}$ haloes, and by $Ω_{\rm m}$ for $M\sim 10^{15} \, \mathrm{M_\odot}\, h^{-1}$. The compact representation learnt by our model can inform the design of emulator training sets to achieve high emulator accuracy with fewer simulations.

astro-ph.CO↗

arXiv2024

A deep learning model for the density profiles of subhaloes in IllustrisTNG

We present a machine-learning-based model for the total density profiles of subhaloes with masses $M \gtrsim 7\times 10^8\,h^{-1}{\rm M}_\odot$ in the IllustrisTNG100 simulation. The model is based on an interpretable variational encoder (IVE) which returns the independent factors of variation in the density profiles within a low-dimensional representation, as well as the predictions for the density profiles themselves. The IVE returns accurate and unbiased predictions on all radial ranges, including the outer region profile where the subhaloes experience tidal stripping; here its fit accuracy exceeds that of the commonly used Einasto profile. The IVE discovers three independent degrees of freedom in the profiles, which can be interpreted in terms of the formation history of the subhaloes. In addition to the two parameters controlling the normalization and inner shape of the profile, the IVE discovers a third parameter that accounts for the impact of tidal stripping onto the subhalo outer profile; this parameter is sensitive to the mass loss experienced by the subhalo after its infall onto its parent halo. Baryonic physics in the IllustrisTNG galaxy formation model does not impact the number of degrees of freedom identified in the profile compared to the pure dark matter expectations, nor their physical interpretation. Our newly proposed profile fit can be used in strong lensing analyses or other observational studies which aim to constrain cosmology from small-scale structures.

astro-ph.GA↗

arXiv2024

Deep learning insights into cosmological structure formation

The evolution of linear initial conditions present in the early universe into extended halos of dark matter at late times can be computed using cosmological simulations. However, a theoretical understanding of this complex process remains elusive; in particular, the role of anisotropic information in the initial conditions in establishing the final mass of dark matter halos remains a long-standing puzzle. Here, we build a deep learning framework to investigate this question. We train a three-dimensional convolutional neural network (CNN) to predict the mass of dark matter halos from the initial conditions, and quantify in full generality the amounts of information in the isotropic and anisotropic aspects of the initial density field about final halo masses. We find that anisotropies add a small, albeit statistically significant amount of information over that contained within spherical averages of the density field about final halo mass. However, the overall scatter in the final mass predictions does not change qualitatively with this additional information, only decreasing from 0.9 dex to 0.7 dex. Given such a small improvement, our results demonstrate that isotropic aspects of the initial density field essentially saturate the relevant information about final halo mass. Therefore, instead of searching for information directly encoded in initial conditions anisotropies, a more promising route to accurate, fast halo mass predictions is to add approximate dynamical information based e.g. on perturbation theory. More broadly, our results indicate that deep learning frameworks can provide a powerful tool for extracting physical insight into cosmological structure formation.

astro-ph.CO↗

arXiv2024

Explaining dark matter halo density profiles with neural networks

We use explainable neural networks to connect the evolutionary history of dark matter halos with their density profiles. The network captures independent factors of variation in the density profiles within a low-dimensional representation, which we physically interpret using mutual information. Without any prior knowledge of the halos' evolution, the network recovers the known relation between the early time assembly and the inner profile, and discovers that the profile beyond the virial radius is described by a single parameter capturing the most recent mass accretion rate. The results illustrate the potential for machine-assisted scientific discovery in complicated astrophysical datasets.

astro-ph.CO↗

arXiv2023

Halo assembly bias from a deep learning model of halo formation

We build a deep learning framework that connects the local formation process of dark matter halos to the halo bias. We train a convolutional neural network (CNN) to predict the final mass and concentration of dark matter halos from the initial conditions. The CNN is then used as a surrogate model to derive the response of the halos' mass and concentration to long-wavelength perturbations in the initial conditions, and consequently the halo bias parameters following the "response bias" definition. The CNN correctly predicts how the local properties of dark matter halos respond to changes in the large-scale environment, despite no explicit knowledge of halo bias being provided during training. We show that the CNN recovers the known trends for the linear and second-order density bias parameters $b_1$ and $b_2$, as well as for the local primordial non-Gaussianity linear bias parameter $b_ϕ$. The expected secondary assembly bias dependence on halo concentration is also recovered by the CNN: at fixed mass, halo concentration has only a mild impact on $b_1$, but a strong impact on $b_ϕ$. Our framework opens a new window for discovering which physical aspects of the halo's Lagrangian patch determine assembly bias, which in turn can inform physical models of halo formation and bias.

astro-ph.CO↗

arXiv2023

A robust estimator of mutual information for deep learning interpretability

We develop the use of mutual information (MI), a well-established metric in information theory, to interpret the inner workings of deep learning models. To accurately estimate MI from a finite number of samples, we present GMM-MI (pronounced $``$Jimmie$"$), an algorithm based on Gaussian mixture models that can be applied to both discrete and continuous settings. GMM-MI is computationally efficient, robust to the choice of hyperparameters and provides the uncertainty on the MI estimate due to the finite sample size. We extensively validate GMM-MI on toy data for which the ground truth MI is known, comparing its performance against established mutual information estimators. We then demonstrate the use of our MI estimator in the context of representation learning, working with synthetic data and physical datasets describing highly non-linear processes. We train deep learning models to encode high-dimensional data within a meaningful compressed (latent) representation, and use GMM-MI to quantify both the level of disentanglement between the latent variables, and their association with relevant physical quantities, thus unlocking the interpretability of the latent representation. We make GMM-MI publicly available.

physics.data-an↗

arXiv2022

Insights into the origin of halo mass profiles from machine learning

The mass distribution of dark matter haloes is the result of the hierarchical growth of initial density perturbations through mass accretion and mergers. We use an interpretable machine-learning framework to provide physical insights into the origin of the spherically-averaged mass profile of dark matter haloes. We train a gradient-boosted-trees algorithm to predict the final mass profiles of cluster-sized haloes, and measure the importance of the different inputs provided to the algorithm. We find two primary scales in the initial conditions (ICs) that impact the final mass profile: the density at approximately the scale of the haloes' Lagrangian patch $R_L$ ($R\sim 0.7\, R_L$) and that in the large-scale environment ($R\sim 1.7~R_L$). The model also identifies three primary time-scales in the halo assembly history that affect the final profile: (i) the formation time of the virialized, collapsed material inside the halo, (ii) the dynamical time, which captures the dynamically unrelaxed, infalling component of the halo over its first orbit, (iii) a third, most recent time-scale, which captures the impact on the outer profile of recent massive merger events. While the inner profile retains memory of the ICs, this information alone is insufficient to yield accurate predictions for the outer profile. As we add information about the haloes' mass accretion history, we find a significant improvement in the predicted profiles at all radii. Our machine-learning framework provides novel insights into the role of the ICs and the mass assembly history in determining the final mass profile of cluster-sized haloes.

astro-ph.CO↗

arXiv2022

Discovering the building blocks of dark matter halo density profiles with neural networks

The density profiles of dark matter halos are typically modeled using empirical formulae fitted to the density profiles of relaxed halo populations. We present a neural network model that is trained to learn the mapping from the raw density field containing each halo to the dark matter density profile. We show that the model recovers the widely-used Navarro-Frenk-White (NFW) profile out to the virial radius, and can additionally describe the variability in the outer profile of the halos. The neural network architecture consists of a supervised encoder-decoder framework, which first compresses the density inputs into a low-dimensional latent representation, and then outputs $ρ(r)$ for any desired value of radius $r$. The latent representation contains all the information used by the model to predict the density profiles. This allows us to interpret the latent representation by quantifying the mutual information between the representation and the halos' ground-truth density profiles. A two-dimensional representation is sufficient to accurately model the density profiles up to the virial radius; however, a three-dimensional representation is required to describe the outer profiles beyond the virial radius. The additional dimension in the representation contains information about the infalling material in the outer profiles of dark matter halos, thus discovering the splashback boundary of halos without prior knowledge of the halos' dynamical history.

astro-ph.CO↗

arXiv2021

The causal effect of environment on halo mass and concentration

Understanding the impact of environment on the formation and evolution of dark matter halos and galaxies is a crucial open problem. Studying statistical correlations in large simulated populations sheds some light on these impacts, but the causal effect of an environment on individual objects is harder to pinpoint. Addressing this, we present a new method for resimulating a single dark matter halo in multiple large-scale environments. In the initial conditions, we 'splice' (i.e. insert) the Lagrangian region of a halo into different Gaussian random fields, while enforcing consistency with the statistical properties of $Λ$CDM. Applying this technique, we demonstrate that the mass of halos is primarily determined by the density structure inside their Lagrangian patches, while the halos' concentration is more strongly affected by environment. The splicing approach will also allow us to study, for example, the impact of the cosmic web on accretion processes and galaxy quenching.

astro-ph.CO↗

arXiv2019

Algorithms and Statistical Models for Scientific Discovery in the Petabyte Era

The field of astronomy has arrived at a turning point in terms of size and complexity of both datasets and scientific collaboration. Commensurately, algorithms and statistical models have begun to adapt --- e.g., via the onset of artificial intelligence --- which itself presents new challenges and opportunities for growth. This white paper aims to offer guidance and ideas for how we can evolve our technical and collaborative frameworks to promote efficient algorithmic development and take advantage of opportunities for scientific discovery in the petabyte era. We discuss challenges for discovery in large and complex data sets; challenges and requirements for the next stage of development of statistical methodologies and algorithmic tool sets; how we might change our paradigms of collaboration and education; and the ethical implications of scientists' contributions to widely applicable algorithms and computational modeling. We start with six distinct recommendations that are supported by the commentary following them. This white paper is related to a larger corpus of effort that has taken place within and around the Petabytes to Science Workshops (https://petabytestoscience.github.io/).

astro-ph.IM↗

arXiv2019

An interpretable machine learning framework for dark matter halo formation

We present a generalization of our recently proposed machine learning framework, aiming to provide new physical insights into dark matter halo formation. We investigate the impact of the initial density and tidal shear fields on the formation of haloes over the mass range $11.4 \leq \log(M/M_{\odot}) \leq 13.4$. The algorithm is trained on an N-body simulation to infer the final mass of the halo to which each dark matter particle will later belong. We then quantify the difference in the predictive accuracy between machine learning models using a metric based on the Kullback-Leibler divergence. We first train the algorithm with information about the density contrast in the particles' local environment. The addition of tidal shear information does not yield an improved halo collapse model over one based on density information alone; the difference in their predictive performance is consistent with the statistical uncertainty of the density-only based model. This implies that our machine learning setup does not identify any significant role for the tidal shear in determining halo masses. This result is confirmed as we verify the ability of the initial conditions-to-halo mass mapping learnt from one simulation to generalize to independent simulations. Our work illustrates the broader potential of developing interpretable machine learning frameworks to gain physical understanding of non-linear large-scale structure formation.

astro-ph.CO↗