arXiv subjects

Michelle Ntampaka

Publications and source records attributed to Michelle Ntampaka.

At least 19 recordsLinked to original sources

arXiv2026

How to Craft the Right Language AI Policy For Your Research Group (Some Assembly Required)

Language AI is rapidly becoming part of the astronomy research ecosystem, prompting research teams to develop policies governing its use. But resources and advice for AI adoption assume that all research groups share the same goals and values. This paper lays out an argument for why there is no single "correct" AI policy for astronomy research groups. Instead, we introduce four research laboratory archetypes with competing research priorities, and we use them to explore how a small laboratory or research group's priorities shape decisions about AI's impact on research productivity, scientist development, scientific integrity, and data governance. Rather than prescribing a universal set of rules, this paper provides a framework for aligning AI policies with a group's scientific values and mission. The objective is to help research leaders decide how Language AI should be used within their particular research environment.

astro-ph.IM↗

arXiv2026

LoVoCCS. III. Third Generation Pipeline & The Hercules Supercluster

The Local Volume Complete Cluster Survey (LoVoCCS) is a volume-complete survey of over one-hundred nearby ($0.03 < z < 0.12$), X-ray luminous ($L_{500} > 10^{44} \text{ erg s}^{-1}$) galaxy clusters in the southern sky. Observations for the survey concluded in December 2025, reaching Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST) Year 1-2 depth in each field and providing observations with $\lesssim 1"$ seeing for weak lensing science. In this paper, we present the latest pipeline for reducing observations using the third-generation of the LSST Science Pipelines. We use recent observations of the Hercules Supercluster to validate the pipeline's data-products and conduct an extensive multi-plane weak-lensing analysis of a $\sim 16 \text{ deg}^2$ complex covering Abell 2147, 2151, 2152, and several additional structures. We confirm that the dynamical mass of the complex is biased due to the dynamical state of Abell 2147, which is consistent with being $\sim 0.2-0.4 \text{ Gyr}$ out-of periapsis, and estimate that the total mass of the supercluster is $8.9^{+1.7}_{-1.4} \times 10^{14}~M_{\odot}$.

astro-ph.GA↗

arXiv2025

Estimating the triaxiality of massive clusters from 2D observables in MillenniumTNG with machine learning

Properties of massive galaxy clusters, such as mass abundance and concentration, are sensitive to cosmology, making cluster statistics a powerful tool for cosmological studies. However, favoring a more simplified, spherically symmetric model for galaxy clusters can lead to biases in the estimates of cluster properties. In this work, we present a deep-learning approach for estimating the triaxiality and orientations of massive galaxy clusters (those with masses $\gtrsim 10^{14}\,M_\odot h^{-1}$) from 2D observables. We utilize the flagship hydrodynamical volume of the suite of cosmological-hydrodynamical MillenniumTNG (MTNG) simulations as our ground truth. Our model combines the feature extracting power of a convolutional neural network (CNN) and the message passing power of a graph neural network (GNN) in a multi-modal, fusion network. Our model is able to extract 3D geometry information from 2D idealized cluster multi-wavelength images (soft X-ray, medium X-ray, hard X-ray and tSZ effect) and mathematical graph representations of 2D cluster member observables (line-of-sight radial velocities, 2D projected positions and V-band luminosities). Our network improves cluster geometry estimation in MTNG by $30\%$ compared to assuming spherical symmetry. We report an $R^2 = 0.85$ regression score for estimating the major axis length of triaxial clusters and correctly classifying $71\%$ of prolate clusters with elongated orientations along our line-of-sight.

astro-ph.CO↗

arXiv2025

The Importance of Being Adaptable: An Exploration of the Power and Limitations of Domain Adaptation for Simulation-Based Inference with Galaxy Clusters

The application of deep machine learning methods in astronomy has exploded in the last decade, with new models showing remarkably improved performance on benchmark tasks. Not nearly enough attention is given to understanding the models' robustness, especially when the test data are systematically different from the training data, or "out of domain." Domain shift poses a significant challenge for simulation-based inference, where models are trained on simulated data but applied to real observational data. In this paper, we explore domain shift and test domain adaptation methods for a specific scientific case: simulation-based inference for estimating galaxy cluster masses from X-ray profiles. We build datasets to mimic simulation-based inference: a training set from the Magneticum simulation, a scatter-augmented training set to capture uncertainties in scaling relations, and a test set derived from the IllustrisTNG simulation. We demonstrate that the Test Set is out of domain in subtle ways that would be difficult to detect without careful analysis. We apply three deep learning methods: a standard neural network (NN), a neural network trained on the scatter-augmented input catalogs, and a Deep Reconstruction-Regression Network (DRRN), a semi-supervised deep model engineered to address domain shift. Although the NN improves results by 17% in the Training Data, it performs 40% worse on the out-of-domain Test Set. Surprisingly, the Scatter-Augmented Neural Network (SANN) performs similarly. While the DRRN is successful in mapping the training and Test Data onto the same latent space, it consistently underperforms compared to a straightforward Yx scaling relation. These results serve as a warning that simulation-based inference must be handled with extreme care, as subtle differences between training simulations and observational data can lead to unforeseen biases creeping into the results.

astro-ph.IM↗

arXiv2025

Reconstructing Galaxy Cluster Mass Maps using Score-based Generative Modeling

We present a novel approach to reconstruct gas and dark matter projected density maps of galaxy clusters using score-based generative modeling. Our diffusion model takes in mock SZ and X-ray images as conditional inputs, and generates realizations of corresponding gas and dark matter maps by sampling from a learned data posterior. We train and validate the performance of our model by using mock data from a cosmological simulation. The model accurately reconstructs both the mean and spread of the radial density profiles in the spatial domain, indicating that the model is able to distinguish between clusters of different mass sizes. In the spectral domain, the model achieves close-to-unity values for the bias and cross-correlation coefficients, indicating that the model can accurately probe cluster structures on both large and small scales. Our experiments demonstrate the ability of score models to learn a strong, nonlinear, and unbiased mapping between input observables and fundamental density distributions of galaxy clusters. These diffusion models can be further fine-tuned and generalized to not only take in additional observables as inputs, but also real observations and predict unknown density distributions of galaxy clusters.

astro-ph.CO↗

arXiv2025

Galaxy cluster characterization with machine learning techniques

We present an analysis of the X-ray properties of the galaxy cluster population in the z=0 snapshot of the IllustrisTNG simulations, utilizing machine learning techniques to perform clustering and regression tasks. We examine five properties of the hot gas (the central cooling time, the central electron density, the central entropy excess, the concentration parameter, and the cuspiness) which are commonly used as classification metrics to identify cool core (CC), weak cool core (WCC) and non cool core (NCC) clusters of galaxies. Using mock Chandra X-ray images as inputs, we first explore an unsupervised clustering scheme to see how the resulting groups correlate with the CC/WCC/NCC classification based on the different criteria. We observe that the groups replicate almost exactly the separation of the galaxy cluster images when classifying them based on the concentration parameter. We then move on to a regression task, utilizing a ResNet model to predict the value of all five properties. The network is able to achieve a mean percentage error of 1.8% for the central cooling time, and a balanced accuracy of 0.83 on the concentration parameter, making them the best-performing metrics. Finally, we use simulation-based inference (SBI) to extract posterior distributions for the network predictions. Our neural network simultaneously predicts all five classification metrics using only mock Chandra X-ray images. This study demonstrates that machine learning is a viable approach for analyzing and classifying the large galaxy cluster datasets that will soon become available through current and upcoming X-ray surveys, such as eROSITA.

astro-ph.GA↗

arXiv2024

A Multi-Wavelength Technique for Estimating Galaxy Cluster Mass Accretion Rates

The mass accretion rate of galaxy clusters is a key factor in determining their structure, but a reliable observational tracer has yet to be established. We present a state-of-the-art machine learning model for constraining the mass accretion rate of galaxy clusters from only X-ray and thermal Sunyaev-Zeldovich observations. Using idealized mock observations of galaxy clusters from the MillenniumTNG simulation, we train a machine learning model to estimate the mass accretion rate. The model constrains 68% of the mass accretion rates of the clusters in our dataset to within 33% of the true value without significant bias, a ~58% reduction in the scatter over existing constraints. We demonstrate that the model uses information from both radial surface brightness density profiles and asymmetries.

astro-ph.CO↗

arXiv2024

Hydroxyl Lines and Moonlight: a High Spectral Resolution Investigation of NIR skylines from Maunakea to guide NIR spectroscopic surveys

Subtracting the changing sky contribution from the near-infrared (NIR) spectra of faint astronomical objects is challenging and crucial to a wide range of science cases such as estimating the velocity dispersions of dwarf galaxies, studying the gas dynamics in faint galaxies, measuring accurate redshifts, and any spectroscopic studies of faint targets. Since the sky background varies with time and location, NIR spectral observations, especially those employing fiber spectrometers and targeting extended sources, require frequent sky-only observations for calibration. However, sky subtraction can be optimized with sufficient a priori knowledge of the sky's variability. In this work, we explore how to optimize sky subtraction by analyzing 1075 high-resolution NIR spectra from the CFHT's SPIRou on Maunakea, and we estimate the variability of 481 hydroxyl (OH) lines. These spectra were collected during two sets of three nights dedicated to obtaining sky observations every five and a half minutes. During the first set, we observed how the Moon affects the NIR, which has not been accurately measured at these wavelengths. We suggest accounting for the Moon contribution at separation distances less than 10 degrees when 1) reconstructing the sky using principal component analysis 2) observing targets at Y JHK mags fainter than ~15 and 3) attempting a sky subtraction better than 1%. We also identified 126 spectral doublets, or OH lines that split into at least two components, at SPIRou's resolution. In addition, we used Lomb-Scargle Periodograms and Gaussian process regression to estimate that most OH lines vary on similar timescales, which provides a valuable input for IR spectroscopic survey strategies. The data and code developed for this study are publicly available.

astro-ph.IM↗

arXiv2024

LoVoCCS. II. Weak Lensing Mass Distributions, Red-Sequence Galaxy Distributions, and Their Alignment with the Brightest Cluster Galaxy in 58 Nearby X-ray-Luminous Galaxy Clusters

The Local Volume Complete Cluster Survey (LoVoCCS) is an on-going program to observe nearly a hundred low-redshift X-ray-luminous galaxy clusters (redshifts $0.03 10^{44}$ erg/s) with the Dark Energy Camera (DECam), capturing data in $u,g,r,i,z$ bands with a $5σ$ point source depth of approximately 25-26th AB magnitudes. Here, we map the aperture masses in 58 galaxy cluster fields using weak gravitational lensing. These clusters span a variety of dynamical states, from nearly relaxed to merging systems, and approximately half of them have not been subject to detailed weak lensing analysis before. In each cluster field, we analyze the alignment between the 2D mass distribution described by the aperture mass map, the 2D red-sequence (RS) galaxy distribution, and the brightest cluster galaxy (BCG). We find that the orientations of the BCG and the RS distribution are strongly aligned throughout the interiors of the clusters: the median misalignment angle is 19 deg within 2 Mpc. We also observe the alignment between the orientations of the RS distribution and the overall cluster mass distribution (by a median difference of 32 deg within 1 Mpc), although this is constrained by galaxy shape noise and the limitations of our cluster sample size. These types of alignment suggest long-term dynamical evolution within the clusters over cosmic timescales.

astro-ph.CO↗

arXiv2024

Designing an Evaluation Framework for Large Language Models in Astronomy Research

Large Language Models (LLMs) are shifting how scientific research is done. It is imperative to understand how researchers interact with these models and how scientific sub-communities like astronomy might benefit from them. However, there is currently no standard for evaluating the use of LLMs in astronomy. Therefore, we present the experimental design for an evaluation study on how astronomy researchers interact with LLMs. We deploy a Slack chatbot that can answer queries from users via Retrieval-Augmented Generation (RAG); these responses are grounded in astronomy papers from arXiv. We record and anonymize user questions and chatbot answers, user upvotes and downvotes to LLM responses, user feedback to the LLM, and retrieved documents and similarity scores with the query. Our data collection method will enable future dynamic evaluations of LLM tools for astronomy.

astro-ph.IM↗

arXiv2023

Predicting the impact of feedback on matter clustering with machine learning in CAMELS

Extracting information from the total matter power spectrum with the precision needed for upcoming cosmological surveys requires unraveling the complex effects of galaxy formation processes on the distribution of matter. We investigate the impact of baryonic physics on matter clustering at $z=0$ using a library of power spectra from the Cosmology and Astrophysics with MachinE Learning Simulations (CAMELS) project, containing thousands of $(25\,h^{-1}{\rm Mpc})^3$ volume realizations with varying cosmology, initial random field, stellar and AGN feedback strength and sub-grid model implementation methods. We show that baryonic physics affects matter clustering on scales $k \gtrsim 0.4\,h\,\mathrm{Mpc}^{-1}$ and the magnitude of this effect is dependent on the details of the galaxy formation implementation and variations of cosmological and astrophysical parameters. Increasing AGN feedback strength decreases halo baryon fractions and yields stronger suppression of power relative to N-body simulations, while stronger stellar feedback often results in weaker effects by suppressing black hole growth and therefore the impact of AGN feedback. We find a broad correlation between mean baryon fraction of massive halos ($M_{\rm 200c} > 10^{13.5}$\,\Msun) and suppression of matter clustering but with significant scatter compared to previous work owing to wider exploration of feedback parameters and cosmic variance effects. We show that a random forest regressor trained on the baryon content and abundance of halos across the full mass range $10^{10} \leq M_\mathrm{halo}/$\Msun$< 10^{15}$ can predict the effect of galaxy formation on the matter power spectrum on scales $k = 1.0$--20.0\,$h\,\mathrm{Mpc}^{-1}$.

astro-ph.GA↗

arXiv2023

Painting baryons onto N-body simulations of galaxy clusters with image-to-image deep learning

Galaxy cluster mass functions are a function of cosmology, but mass is not a direct observable, and systematic errors abound in all its observable proxies. Mass-free inference can bypass this challenge, but it requires large suites of simulations spanning a range of cosmologies and models for directly observable quantities. In this work, we devise a U-net - an image-to-image machine learning algorithm - to ``paint'' the IllustrisTNG model of baryons onto dark-matter-only simulations of galaxy clusters. Using 761 galaxy clusters with $M_{200c} \gtrsim 10^{14}M_\odot$ from the TNG-300 simulation at $z<1$, we train the algorithm to read in maps of projected dark matter mass and output maps of projected gas density, temperature, and X-ray flux. The models train in under an hour on two GPUs, and then predict baryonic images for $\sim2700$ dark matter maps drawn from the TNG-300 dark-matter-only (DMO) simulation in under two minutes. Despite being trained on individual images, the model reproduces the true scaling relation and scatter for the $M_{DM}-L_X$, as well as the distribution functions of the cluster X-ray luminosity and gas mass. For just one decade in cluster mass, the model reproduces three orders of magnitude in $L_X$. The model is biased slightly high when using dark matter maps from the DMO simulation. The model performs well on inputs from TNG-300-2, whose mass resolution is 8 times coarser; further degrading the resolution biases the predicted luminosity function high. We conclude that U-net-based baryon painting is a promising technique to build large simulated cluster catalogs which can be used to improve cluster cosmology by combining existing full-physics and large $N$-body simulations.

astro-ph.CO↗

arXiv2023

Benchmarks and Explanations for Deep Learning Estimates of X-ray Galaxy Cluster Masses

We evaluate the effectiveness of deep learning (DL) models for reconstructing the masses of galaxy clusters using X-ray photometry data from next-generation surveys. We establish these constraints using a catalogue of realistic mock eROSITA X-ray observations which use hydrodynamical simulations to model realistic cluster morphology, background emission, telescope response, and AGN sources. Using bolometric X-ray photon maps as input, DL models achieve a predictive mass scatter of $σ_{\ln M_\mathrm{500c}} = 17.8\%$, a factor of two improvements on scalar observables such as richness $N_\mathrm{gal}$, 1D velocity dispersion $σ_\mathrm{v,1D}$, and photon count $N_\mathrm{phot}$ as well as a $32\%$ improvement upon idealised, volume-integrated measurements of the bolometric X-ray luminosity $L_X$. We then show that extending this model to handle multichannel X-ray photon maps, separated in low, medium, and high energy bands, further reduces the mass scatter to $16.2\%$. We also tested a multimodal DL model incorporating both dynamical and X-ray cluster probes and achieved marginal gains at a mass scatter of $15.9\%$. Finally, we conduct a quantitative interpretability study of our DL models and find that they greatly down-weight the importance of pixels in the centres of clusters and at the location of AGN sources, validating previous claims of DL modelling improvements and suggesting practical and theoretical benefits for using DL in X-ray mass inference.

astro-ph.CO↗

arXiv2022

A Machine Learning Approach to Enhancing eROSITA Observations

The eROSITA X-ray telescope, launched in 2019, is predicted to observe roughly 100,000 galaxy clusters. Follow-up observations of these clusters from Chandra, for example, will be needed to resolve outstanding questions about galaxy cluster physics. Deep Chandra cluster observations are expensive and follow-up of every eROSITA cluster is infeasible, therefore, objects chosen for follow-up must be chosen with care. To address this, we have developed an algorithm for predicting longer duration, background-free observations based on mock eROSITA observations. We make use of the hydrodynamic cosmological simulation Magneticum, have simulated eROSITA instrument conditions using SIXTE, and have applied a novel convolutional neural network to output a deep Chandra-like "super observation" of each cluster in our simulation sample. Any follow-up merit assessment tool should be designed with a specific use case in mind; our model produces observations that accurately and precisely reproduce the cluster morphology, which is a critical ingredient for determining cluster dynamical state and core type. Our model will advance our understanding of galaxy clusters by improving follow-up selection and demonstrates that image-to-image deep learning algorithms are a viable method for simulating realistic follow-up observations.

astro-ph.CO↗

arXiv2022

The Dynamical Mass of the Coma Cluster from Deep Learning

In 1933, Fritz Zwicky's famous investigations of the mass of the Coma cluster led him to infer the existence of dark matter \cite{1933AcHPh...6..110Z}. His fundamental discoveries have proven to be foundational to modern cosmology; as we now know such dark matter makes up 85\% of the matter and 25\% of the mass-energy content in the universe. Galaxy clusters like Coma are massive, complex systems of dark matter in addition to hot ionized gas and thousands of galaxies, and serve as excellent probes of the dark matter distribution. However, empirical studies show that the total mass of such systems remains elusive and difficult to precisely constrain. Here, we present new estimates for the dynamical mass of the Coma cluster based on Bayesian deep learning methodologies developed in recent years. Using our novel data-driven approach, we predict Coma's $\mthc$ mass to be $10^{15.10 \pm 0.15}\ \hmsun$ within a radius of $1.78 \pm 0.03\ h^{-1}\mathrm{Mpc}$ of its center. We show that our predictions are rigorous across multiple training datasets and statistically consistent with historical estimates of Coma's mass. This measurement reinforces our understanding of the dynamical state of the Coma cluster and advances rigorous analyses and verification methods for empirical applications of machine learning in astronomy.

astro-ph.CO↗

arXiv2022

A Referee Primer for Early Career Astronomers

Refereeing is a crucial component of publishing astronomical research, but few professional astronomers receive formal training on how to effectively referee a manuscript. In this article, we lay out considerations and best practices for referees. This document is intended as a tool for early career researchers to develop a fair, effective, and efficient approach to refereeing.

astro-ph.IM↗

arXiv2022

R2-D2: Roman and Rubin -- From Data to Discovery

The NASA Nancy Grace Roman Space Telescope (Roman) and the Vera C. Rubin Observatory Legacy Survey of Space and Time (Rubin), will transform our view of the wide-field sky, with similar sensitivities, but complementary in wavelength, spatial resolution, and time domain coverage. Here we present findings from the AURA Roman+Rubin Synergy Working group, charged by the STScI and NOIRLab Directors to identify frontier science questions in General Astrophysics, beyond the well-covered areas of Dark Energy and Cosmology, that can be uniquely addressed with Roman and Rubin synergies in observing strategy, data products and archiving, joint analysis, and community engagement. This analysis was conducted with input from the community in the form of brief (1-2 paragraph) "science pitches" (see Appendix), and testimony from "outside experts" (included as co-authors). We identify a rich and broad landscape of potential discoveries catalyzed by the combination of exceptional quality and quantity of Roman and Rubin data, and summarize implementation requirements that would facilitate this bounty of additional science with coordination of survey fields, joint coverage of the Galactic plane, bulge, and ecliptic, expansion of General Investigator and Target of Opportunity observing modes, co-location of Roman and Rubin data, and timely distribution of data, transient alerts, catalogs, value-added joint analysis products, and simulations to the broad astronomical community.

astro-ph.IM↗

arXiv2022

Emulating Sunyaev-Zeldovich Images of Galaxy Clusters using Auto-Encoders

We develop a machine learning algorithm that generates high-resolution thermal Sunyaev-Zeldovich (SZ) maps of novel galaxy clusters given only halo mass and mass accretion rate. The algorithm uses a conditional variational autoencoder (CVAE) in the form of a convolutional neural network and is trained with SZ maps generated from the IllustrisTNG simulation. Our method can reproduce many of the details of galaxy clusters that analytical models usually lack, such as internal structure and aspherical distribution of gas created by mergers, while achieving the same computational feasibility, allowing us to generate mock SZ maps for over $10^5$ clusters in 30 seconds on a laptop. We show that the model is capable of generating novel clusters (i.e. not found in the training set) and that the model accurately reproduces the effects of mass and mass accretion rate on the SZ images, such as scatter, asymmetry, and concentration, in addition to modeling merging sub-clusters. This work demonstrates the viability of machine-learning--based methods for producing the number of realistic, high-resolution maps of galaxy clusters necessary to achieve statistical constraints from future SZ surveys.

astro-ph.CO↗