arXiv subjects

François Lanusse

Publications and source records attributed to François Lanusse.

At least 19 recordsLinked to original sources

arXiv2026

The Cross-Survey Decade: A Call to Action

By 2027, three flagship wide-field surveys will be operating simultaneously from ground and space, observing overlapping sky and representing more than $6 billion in US and European public investment. Together they will produce overlapping petabyte-scale datasets across thousands of square degrees. This is a different class of challenge: the observations are no longer the bottleneck; realizing their joint scientific return now depends on shared computational infrastructure and coordination. Decades of community studies show that combining these datasets does more than improve precision. For science ranging from weak lensing to transient discovery and Galactic-plane astronomy, joint processing and analysis can unlock capabilities no single survey provides alone. Yet the required infrastructure -- joint pixel-level processing, cross-calibration and validation, interoperable data access, and the people to build and sustain it -- falls outside any single mission or institution's mandate. We issue a call to action for cross-survey science infrastructure, built around four pillars: (1) joint pixel-level processing and validation; (2) an AI-ready data substrate for scientific foundation models; (3) standardized, interoperable data access across surveys, democratizing participation in astrophysical discovery; and (4) dedicated personnel and career pathways. We outline concrete steps for policymakers, agencies, observatories, universities, the research community, and philanthropy, and argue that the moment to act is now, while foundational technical choices can still be aligned at a fraction of the cost of reconciling them later.

astro-ph.IM↗

arXiv2026

Towards Practical Field-Level Inference for Weak Lensing

Nonlinear structure growth generates higher-order correlations and morphological features in the cosmic density field that cannot be fully characterized by two-point statistics. Upcoming surveys will measure these features with greater precision, making it essential to develop methods capable of extracting as much cosmological information as possible from them. Field-level inference (FLI) is one such approach, in which cosmological parameters are constrained by comparing observed maps to forward-modeled maps, either directly or through learned summaries that retain map-level information. In this work, we compare FLI with power-spectrum-based inference using the same forward-modeling pipeline for generating weak lensing maps, with the goal of quantifying the gain from map-level analysis relative to two-point statistics. We perform this comparison with both implicit and explicit inference methods, using 8-million-parameter forward models based on Lagrangian perturbation theory and particle-mesh (PM) N-body evolution. The two FLI approaches yield closely consistent posteriors; this agreement, together with coverage tests confirming the calibration of the implicit analyses, gives us confidence in the recovered field-level constraints. Relative to the power-spectrum-based analyses, these results show significant gains in cosmological information, especially when small scales are included in the PM-based forward model. We then discuss the remaining challenges that must be addressed before PM-based explicit FLI can be applied to observational datasets.

astro-ph.CO↗

arXiv2026

Representation Learning for Spatiotemporal Physical Systems

Machine learning approaches to spatiotemporal physical systems have primarily focused on next-frame prediction, with the goal of learning an accurate emulator for the system's evolution in time. However, these emulators are computationally expensive to train and are subject to performance pitfalls, such as compounding errors during autoregressive rollout. In this work, we take a different perspective and look at scientific tasks further downstream of predicting the next frame, such as estimation of a system's governing physical parameters. Accuracy on these tasks offers a uniquely quantifiable glimpse into the physical relevance of the representations of these models. We evaluate the effectiveness of general-purpose self-supervised methods in learning physics-grounded representations that are useful for downstream scientific tasks. Surprisingly, we find that not all methods designed for physical modeling outperform generic self-supervised learning methods on these tasks, and methods that learn in the latent space (e.g., joint embedding predictive architectures, or JEPAs) outperform those optimizing pixel-level prediction objectives. Code is available at https://github.com/helenqu/physical-representation-learning.

cs.LG↗

arXiv2026

Opportunities in AI/ML for the Rubin LSST Dark Energy Science Collaboration

The Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST) will produce unprecedented volumes of heterogeneous astronomical data (images, catalogs, and alerts) that challenge traditional analysis pipelines. The LSST Dark Energy Science Collaboration (DESC) aims to derive robust constraints on dark energy and dark matter from these data, requiring methods that are statistically powerful, scalable, and operationally reliable. Artificial intelligence and machine learning (AI/ML) are already embedded across DESC science workflows, from photometric redshifts and transient classification to weak lensing inference and cosmological simulations. Yet their utility for precision cosmology hinges on trustworthy uncertainty quantification, robustness to covariate shift and model misspecification, and reproducible integration within scientific pipelines. This white paper surveys the current landscape of AI/ML across DESC's primary cosmological probes and cross-cutting analyses, revealing that the same core methodologies and fundamental challenges recur across disparate science cases. Since progress on these cross-cutting challenges would benefit multiple probes simultaneously, we identify key methodological research priorities, including Bayesian inference at scale, physics-informed methods, validation frameworks, and active learning for discovery. With an eye on emerging techniques, we also explore the potential of the latest foundation model methodologies and LLM-driven agentic AI systems to reshape DESC workflows, provided their deployment is coupled with rigorous evaluation and governance. Finally, we discuss critical software, computing, data infrastructure, and human capital requirements for the successful deployment of these new methodologies, and consider associated risks and opportunities for broader coordination with external actors.

astro-ph.IM↗

arXiv2025

Benchmarking field-level cosmological inference from galaxy redshift surveys

Field-level inference has emerged as a promising framework to fully harness the cosmological information encoded in next-generation galaxy surveys. It involves performing Bayesian inference to jointly estimate the cosmological parameters and the initial conditions of the cosmic field, directly from the observed galaxy density field. Yet, the scalability and efficiency of sampling algorithms for field-level inference of large-scale surveys remain unclear. To address this, we introduce a standardized benchmark using a fast and differentiable simulator for the galaxy density field based on $\texttt{JaxPM}$. We evaluate a range of sampling methods, including standard Hamiltonian Monte Carlo (HMC), No-U-Turn Sampler (NUTS) without and within a Gibbs scheme, and both adjusted and unadjusted microcanonical samplers (MAMS and MCLMC). These methods are compared based on their efficiency, in particular the number of model evaluations required per effective posterior sample. Our findings emphasize the importance of carefully preconditioning latent variables and demonstrate the significant advantage of (unadjusted) MCLMC for scaling to $\geq 10^6$-dimensional problems. We find that MCLMC outperforms adjusted samplers by over an order-of-magnitude, with a mild scaling with the dimension of our inference problem. This benchmark, along with the associated publicly available code, is intended to serve as a basis for the evaluation of future field-level sampling strategies. The code is readily open-sourced at https://github.com/hsimonfroy/benchmark-field-level

astro-ph.CO↗

arXiv2025

Predicting partially observable dynamical systems via diffusion models with a multiscale inference scheme

Conditional diffusion models provide a natural framework for probabilistic prediction of dynamical systems and have been successfully applied to fluid dynamics and weather prediction. However, in many settings, the available information at a given time represents only a small fraction of what is needed to predict future states, either due to measurement uncertainty or because only a small fraction of the state can be observed. This is true for example in solar physics, where we can observe the Sun's surface and atmosphere, but its evolution is driven by internal processes for which we lack direct measurements. In this paper, we tackle the probabilistic prediction of partially observable, long-memory dynamical systems, with applications to solar dynamics and the evolution of active regions. We show that standard inference schemes, such as autoregressive rollouts, fail to capture long-range dependencies in the data, largely because they do not integrate past information effectively. To overcome this, we propose a multiscale inference scheme for diffusion models, tailored to physical processes. Our method generates trajectories that are temporally fine-grained near the present and coarser as we move farther away, which enables capturing long-range temporal dependencies without increasing computational cost. When integrated into a diffusion model, we show that our inference scheme significantly reduces the bias of the predicted distributions and improves rollout stability.

cs.LG↗

arXiv2025

Learning Diffusion Priors from Observations by Expectation Maximization

Diffusion models recently proved to be remarkable priors for Bayesian inverse problems. However, training these models typically requires access to large amounts of clean data, which could prove difficult in some settings. In this work, we present DiEM, a novel method based on the expectation-maximization algorithm for training diffusion models from incomplete and noisy observations only. Unlike previous works, DiEM leads to proper diffusion models, which is crucial for downstream tasks. As part of our methods, we propose and motivate an improved posterior sampling scheme for unconditional diffusion models. We present empirical evidence supporting the effectiveness of our approach.

cs.LG↗

arXiv2025

Lost in Latent Space: An Empirical Study of Latent Diffusion Models for Physics Emulation

The steep computational cost of diffusion models at inference hinders their use as fast physics emulators. In the context of image and video generation, this computational drawback has been addressed by generating in the latent space of an autoencoder instead of the pixel space. In this work, we investigate whether a similar strategy can be effectively applied to the emulation of dynamical systems and at what cost. We find that the accuracy of latent-space emulation is surprisingly robust to a wide range of compression rates (up to 1000x). We also show that diffusion-based emulators are consistently more accurate than non-generative counterparts and compensate for uncertainty in their predictions with greater diversity. Finally, we cover practical design choices, spanning from architectures to optimizers, that we found critical to train latent-space emulators.

cs.LG↗

arXiv2025

Hybrid Physical-Neural Simulator for Fast Cosmological Hydrodynamics

Cosmological field-level inference requires differentiable forward models that solve the challenging dynamics of gas and dark matter under hydrodynamics and gravity. We propose a hybrid approach where gravitational forces are computed using a differentiable particle-mesh solver, while the hydrodynamics are parametrized by a neural network that maps local quantities to an effective pressure field. We demonstrate that our method improves upon alternative approaches, such as an Enthalpy Gradient Descent baseline, both at the field and summary-statistic level. The approach is furthermore highly data efficient, with a single reference simulation of cosmological structure formation being sufficient to constrain the neural pressure model. This opens the door for future applications where the model is fit directly to observational data, rather than a training set of simulations.

astro-ph.CO↗

arXiv2025

Simulation-Based Inference Benchmark for Weak Lensing Cosmology

Standard cosmological analysis, which relies on two-point statistics, fails to extract the full information of the data. This limits our ability to constrain with precision cosmological parameters. Thus, recent years have seen a paradigm shift from analytical likelihood-based to simulation-based inference. However, such methods require a large number of costly simulations. We focus on full-field inference, considered the optimal form of inference. Our objective is to benchmark several ways of conducting full-field inference to gain insight into the number of simulations required for each method. We make a distinction between explicit and implicit full-field inference. Moreover, as it is crucial for explicit full-field inference to use a differentiable forward model, we aim to discuss the advantages of having this property for the implicit approach. We use the sbi_lens package which provides a fast and differentiable log-normal forward model. This forward model enables us to compare explicit and implicit full-field inference with and without gradient. The former is achieved by sampling the forward model through the No U-Turns sampler. The latter starts by compressing the data into sufficient statistics and uses the Neural Likelihood Estimation algorithm and the one augmented with gradient. We perform a full-field analysis on LSST Y10 like weak lensing simulated mass maps. We show that explicit and implicit full-field inference yield consistent constraints. Explicit inference requires 630 000 simulations with our particular sampler corresponding to 400 independent samples. Implicit inference requires a maximum of 101 000 simulations split into 100 000 simulations to build sufficient statistics (this number is not fine tuned) and 1 000 simulations to perform inference. Additionally, we show that our way of exploiting the gradients does not significantly help implicit inference.

astro-ph.CO↗

arXiv2025

The Cosmological analysis of X-ray cluster surveys VII. Bypassing scaling relations with Lagrangian Deep Learning and Simulation-based inference

Galaxy clusters, the pinnacle of structure formation in our universe, are a powerful cosmological probe. Several approaches have been proposed to express cluster number counts, but all these methods rely on empirical explicit scaling relations that link observed properties to the total cluster mass. These scaling relations are over-parametrised, inducing some degeneracy with cosmology. Moreover, they do not provide a direct handle on the numerous non-gravitational phenomena that affect the physics of the intra-cluster medium. We present a proof-of-concept to model cluster number counts, that bypasses the explicit use of scaling relations. We rather implement the effect of several astrophysical processes to describe the cluster properties. We then evaluate the performances of this modelling for the cosmological inference. We developed an accelerated machine learning baryonic field-emulator, built upon the Lagrangian Deep Learning method and trained on the CAMELS simulations. We then created a pipeline that simulates cluster counts in terms of XMM observable quantities. We finally compare the performances of our model, with that involving scaling relations, for the purpose of cosmological inference based on simulations. Our model correctly reproduces the cluster population from the calibration simulations at the fiducial parameter values, and allows us to constrain feedback mechanisms. The cosmological-inference analyses indicate that our simulation-based model is less degenerate than the approach using scaling relations. This novel approach to model observed cluster number counts from simulations opens interesting perspectives for cluster cosmology. It has the potential to overcome the limitations of the standard approach, provided that the resolution and the volume of the simulations will allow a most realistic implementation of the complex phenomena driving cluster evolution.

astro-ph.CO↗

arXiv2025

DESC DC2 Data Release Note

In preparation for cosmological analyses of the Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST), the LSST Dark Energy Science Collaboration (LSST DESC) has created a 300 deg$^2$ simulated survey as part of an effort called Data Challenge 2 (DC2). The DC2 simulated sky survey, in six optical bands with observations following a reference LSST observing cadence, was processed with the LSST Science Pipelines (19.0.0). In this Note, we describe the public data release of the resulting object catalogs for the coadded images of five years of simulated observations along with associated truth catalogs. We include a brief description of the major features of the available data sets. To enable convenient access to the data products, we have developed a web portal connected to Globus data services. We describe how to access the data and provide example Jupyter Notebooks in Python to aid first interactions with the data. We welcome feedback and questions about the data release via a GitHub repository.

astro-ph.CO↗

arXiv2025

Optimal Neural Summarisation for Full-Field Weak Lensing Cosmological Implicit Inference

Traditionally, weak lensing cosmological surveys have been analyzed using summary statistics motivated by their analytically tractable likelihoods, or by their ability to access higher-order information, at the cost of requiring Simulation-Based Inference (SBI) approaches. While informative, these statistics are neither designed nor guaranteed to be statistically sufficient. With the rise of deep learning, it becomes possible to create summary statistics optimized to extract the full data information. We compare different neural summarization strategies proposed in the weak lensing literature, to assess which loss functions lead to theoretically optimal summary statistics to perform full-field inference. In doing so, we aim to provide guidelines and insights to the community to help guide future neural-based inference analyses. We design an experimental setup to isolate the impact of the loss function used to train neural networks. We have developed the sbi_lens JAX package, which implements an automatically differentiable lognormal wCDM LSST-Y10 weak lensing simulator. The explicit full-field posterior obtained using the Hamiltonian Monte Carlo sampler gives us a ground truth to which to compare different compression strategies. We provide theoretical insight into the loss functions used in the literature and show that some do not necessarily lead to sufficient statistics (e.g. Mean Square Error (MSE)), while those motivated by information theory (e.g. Variational Mutual Information Maximization (VMIM)) can. Our numerical experiments confirm these insights and show, in our simulated wCDM scenario, that the Figure of Merit (FoM) of an analysis using neural summaries optimized under VMIM achieves 100% of the reference Omega_c - sigma_8 full-field FoM, while an analysis using neural summaries trained under MSE achieves only 81% of the same reference FoM.

astro-ph.CO↗

arXiv2024

Differentiable Cosmological Simulation with Adjoint Method

Rapid advances in deep learning have brought not only myriad powerful neural networks, but also breakthroughs that benefit established scientific research. In particular, automatic differentiation (AD) tools and computational accelerators like GPUs have facilitated forward modeling of the Universe with differentiable simulations. Based on analytic or automatic backpropagation, current differentiable cosmological simulations are limited by memory, and thus are subject to a trade-off between time and space/mass resolution, usually sacrificing both. We present a new approach free of such constraints, using the adjoint method and reverse time integration. It enables larger and more accurate forward modeling at the field level, and will improve gradient based optimization and inference. We implement it in an open-source particle-mesh (PM) $N$-body library pmwd (particle-mesh with derivatives). Based on the powerful AD system JAX, pmwd is fully differentiable, and is highly performant on GPUs.

astro-ph.IM↗

arXiv2023

Forecasting the power of Higher Order Weak Lensing Statistics with automatically differentiable simulations

We present the Differentiable Lensing Lightcone (DLL), a fully differentiable physical model designed for being used as a forward model in Bayesian inference algorithms requiring access to derivatives of lensing observables with respect to cosmological parameters. We extend the public FlowPM N-body code, a particle-mesh N-body solver, simulating lensing lightcones and implementing the Born approximation in the Tensorflow framework. Furthermore, DLL is aimed at achieving high accuracy with low computational costs. As such, it integrates a novel Hybrid Physical-Neural parameterisation able to compensate for the small-scale approximations resulting from particle-mesh schemes for cosmological N-body simulations. We validate our simulations in an LSST setting against high-resolution $κ$TNG simulations by comparing both the lensing angular power spectrum and multiscale peak counts. We demonstrate an ability to recover lensing $C_\ell$ up to a 10% accuracy at $\ell=1000$ for sources at redshift 1, with as few as $\sim 0.6$ particles per Mpc/h. As a first use case, we use this tool to investigate the relative constraining power of the angular power spectrum and peak counts statistic in an LSST setting. Such comparisons are typically very costly as they require a large number of simulations, and do not scale well with the increasing number of cosmological parameters. As opposed to forecasts based on finite differences, these statistics can be analytically differentiated with respect to cosmology, or any systematics included in the simulations at the same computational cost of the forward simulation. We find that the peak counts outperform the power spectrum on the cold dark matter parameter $Ω_c$, on the amplitude of density fluctuations $σ_8$, and on the amplitude of the intrinsic alignment signal $A_{IA}$.

astro-ph.IM↗

arXiv2023

JAX-COSMO: An End-to-End Differentiable and GPU Accelerated Cosmology Library

We present jax-cosmo, a library for automatically differentiable cosmological theory calculations. It uses the JAX library, which has created a new coding ecosystem, especially in probabilistic programming. As well as batch acceleration, just-in-time compilation, and automatic optimization of code for different hardware modalities (CPU, GPU, TPU), JAX exposes an automatic differentiation (autodiff) mechanism. Thanks to autodiff, jax-cosmo gives access to the derivatives of cosmological likelihoods with respect to any of their parameters, and thus enables a range of powerful Bayesian inference algorithms, otherwise impractical in cosmology, such as Hamiltonian Monte Carlo and Variational Inference. In its initial release, jax-cosmo implements background evolution, linear and non-linear power spectra (using halofit or the Eisenstein and Hu transfer function), as well as angular power spectra with the Limber approximation for galaxy and weak lensing probes, all differentiable with respect to the cosmological parameters and their other inputs. We illustrate how autodiff can be a game-changer for common tasks involving Fisher matrix computations, or full posterior inference with gradient-based techniques. In particular, we show how Fisher matrices are now fast, exact, no longer require any fine tuning, and are themselves differentiable. Finally, using a Dark Energy Survey Year 1 3x2pt analysis as a benchmark, we demonstrate how jax-cosmo can be combined with Probabilistic Programming Languages to perform posterior inference with state-of-the-art algorithms including a No U-Turn Sampler, Automatic Differentiation Variational Inference,and Neural Transport HMC. We further demonstrate that Normalizing Flows using Neural Transport are a promising methodology for model validation in the early stages of analysis.

astro-ph.CO↗

arXiv2022

pmwd: A Differentiable Cosmological Particle-Mesh $N$-body Library

The formation of the large-scale structure, the evolution and distribution of galaxies, quasars, and dark matter on cosmological scales, requires numerical simulations. Differentiable simulations provide gradients of the cosmological parameters, that can accelerate the extraction of physical information from statistical analyses of observational data. The deep learning revolution has brought not only myriad powerful neural networks, but also breakthroughs including automatic differentiation (AD) tools and computational accelerators like GPUs, facilitating forward modeling of the Universe with differentiable simulations. Because AD needs to save the whole forward evolution history to backpropagate gradients, current differentiable cosmological simulations are limited by memory. Using the adjoint method, with reverse time integration to reconstruct the evolution history, we develop a differentiable cosmological particle-mesh (PM) simulation library pmwd (particle-mesh with derivatives) with a low memory cost. Based on the powerful AD library JAX, pmwd is fully differentiable, and is highly performant on GPUs.

astro-ph.IM↗

arXiv2022

The DAWES review 10: The impact of deep learning for the analysis of galaxy surveys

The amount and complexity of data delivered by modern galaxy surveys has been steadily increasing over the past years. Extracting coherent scientific information from these large and multi-modal data sets remains an open issue and data driven approaches such as deep learning have rapidly emerged as a potentially powerful solution to some long lasting challenges. This enthusiasm is reflected in an unprecedented exponential growth of publications using neural networks. Half a decade after the first published work in astronomy mentioning deep learning, we believe it is timely to review what has been the real impact of this new technology in the field and its potential to solve key challenges raised by the size and complexity of the new datasets. In this review we first aim at summarizing the main applications of deep learning for galaxy surveys that have emerged so far. We then extract the major achievements and lessons learned and highlight key open questions and limitations. Overall, state-of-the art deep learning methods are rapidly adopted by the astronomical community, reflecting a democratization of these methods. We show that the majority of works using deep learning up to date are oriented to computer vision tasks. This is also the domain of application where deep learning has brought the most important breakthroughs so far. We report that the applications are becoming more diverse and deep learning is used for estimating galaxy properties, identifying outliers or constraining the cosmological model. Most of these works remain at the exploratory level. Some common challenges will most likely need to be addressed before moving to the next phase of deployment of deep learning in the processing of future surveys; e.g. uncertainty quantification, interpretability, data labeling and domain shift issues from training with simulations, which constitutes a common practice in astronomy.

astro-ph.IM↗