arXiv subjects

Gilles Louppe

Publications and source records attributed to Gilles Louppe.

At least 19 recordsLinked to original sources

arXiv2026

The Physical Limit of Neural Hypoxia Detection in the Black Sea from Satellite Observations

Coastal hypoxia (O_2 < 63 [mmol / m^3]) threatens ocean health worldwide. On continental shelves, summer stratification prevents bottom oxygen consumed by respiration from being renewed, making monitoring essential to protect vulnerable ecosystems and reduce biodiversity loss. Although satellite observations are increasingly available, their potential to infer subsurface oxygen remains largely unexplored. We frame this as a Bayesian inverse problem relating surface observations to the complete three-dimensional physical and biogeochemical states of the Black Sea. Here, we solve it using a deep generative neural network trained on numerical model outputs that provides a tractable and computationally efficient approximation of the true posterior distribution of sea states. We find that accurate state estimation is limited to the mixed layer, because its homogeneity makes surface conditions representative of subsurface states. During summer, we detect 38% of all hypoxic events shelf-wide with a precision of 47%. Improving the results will likely require longer assimilation windows or subsurface observations.

physics.ao-ph↗

arXiv2026

Improving Semantic Uncertainty Quantification in LVLMs with Semantic Gaussian Processes

Large Vision-Language Models (LVLMs) often produce plausible but unreliable outputs, making robust uncertainty estimation essential. Recent work on semantic uncertainty estimates relies on external models to cluster multiple sampled responses and measure their semantic consistency. However, these clustering methods are often fragile, highly sensitive to minor phrasing variations, and can incorrectly group or separate semantically similar answers, leading to unreliable uncertainty estimates. We propose Semantic Gaussian Process Uncertainty (SGPU), a Bayesian framework that quantifies semantic uncertainty by analyzing the geometric structure of answer embeddings, avoiding brittle clustering. SGPU maps generated answers into a dense semantic space, computes the Gram matrix of their embeddings, and summarizes their semantic configuration via the eigenspectrum. This spectral representation is then fed into a Gaussian Process Classifier that learns to map patterns of semantic consistency to predictive uncertainty, and that can be applied in both black-box and white-box settings. Across six LLMs and LVLMs on eight datasets spanning VQA, image classification, and textual QA, SGPU consistently achieves state-of-the-art calibration (ECE) and discriminative (AUROC, AUARC) performance. We further show that SGPU transfers across models and modalities, indicating that its spectral representation captures general patterns of semantic uncertainty.

cs.CV↗

arXiv2026

An Introduction to Bayesian and Frequentist Simulation-Based Inference with Machine Learning

Simulation-based inference (SBI) with machine learning is an increasingly important tool for solving inverse problems in science and engineering, including parameter inference and the inversion of detector effects. We provide an overview of the Bayesian and frequentist statistical frameworks, describe how machine-learning-based SBI methods, such as neural posterior estimation and neural likelihood estimation, can be used for parameter estimation within these frameworks, and show that the same methods can also be applied to Empirical Bayes or unfolding tasks. We also discuss how to validate inference results and the limitations of SBI with machine learning.

cs.LG↗

arXiv2026

Accounting for Hysteresis and Eddy Currents in Finite Element Simulations of Ferromagnetic Laminated Cores using a Recurrent Neural Network

Incorporating hysteresis and eddy currents into finite element simulations of laminated-core electrical machines is computationally challenging. Resolving the fields inside the laminations at each integration point and at every nonlinear iteration leads to computational costs several orders of magnitude higher than anhysteretic simulations, making such approaches impractical for design applications. Conversely, simplified models accounting only for magnetic saturation are becoming increasingly inadequate as electrical machine topologies and operating conditions grow in complexity. In this context, machine learning surrogate modeling has emerged as a promising alternative, offering efficient and accurate approximations of complex electromagnetic behaviors. In this paper, a recurrent neural network is trained as a surrogate of a laminated-core material model for an isotropic laminated core, and is integrated into realistic two-dimensional magnetodynamic finite element simulations based on a magnetic vector potential formulation. The proposed approach achieves excellent agreement with the reference laminated-core model while limiting the computational cost to about twice that of an anhysteretic simulation. By training the recurrent neural network on a sufficiently diverse set of artificially generated magnetic field sequences designed to mimic those encountered in electrical machine simulations, the proposed approach can be readily applied across a wide range of finite element simulations. Furthermore, the trained surrogate model is provided as a standalone component that can be easily incorporated into existing computational frameworks. It is publicly available at https://gitlab.onelab.info/getdp/lamnet.

cs.CE↗

arXiv2026

Training-Free Bayesian Filtering with Generative Emulators

Bayesian filtering is a well-known problem that aims to estimate plausible states of a dynamical system from observations. Among existing approaches to solve this problem, particle filters are theoretically exact for non-linear dynamics and observations, but suffer from poor scalability in high dimensions. In this work, we show that diffusion-based emulators of dynamical systems can be used to implement, without additional training, an optimal variant of particle filters that has remained largely unexplored due to implementation challenges with classical numerical solvers. Experiments on nonlinear chaotic systems, including atmospheric dynamics, demonstrate that the proposed approach successfully scales particle filtering to high-dimensional settings.

cs.LG↗

arXiv2026

Predict-Project-Renoise: Sampling Diffusion Models under Hard Constraints

Diffusion models cannot enforce hard constraints, yet applications in the physical sciences demand exact satisfaction of conservation laws, boundary conditions, and observational consistency. In this work, we identify a corrector kernel whose unique stationary distribution is the constrained marginal at each noise level, and approximate it by iteratively projecting through the denoiser and renoising via the forward kernel. The resulting Predict-Project-Renoise (PPR) algorithm enables sampling from pretrained diffusion models under hard constraints. Its three components are each necessary: projecting through the denoiser keeps samples close to the data manifold, while renoising and iterating drive samples toward the constrained marginal. On 2D distributions, the Kuramoto-Sivashinsky equation, and global weather forecasting with a $10^8$-dimensional atmospheric model, PPR simultaneously achieves low constraint violations and high distributional fidelity, a combination that existing methods fail to deliver.

cs.LG↗

arXiv2026

Bayesian Inference for Estimating Generation Costs in Electricity Markets

Estimating generation costs from observed electricity market data is essential for market simulation, strategic bidding, and system planning. To that end, we model the relationship between generation costs and production schedules with a latent variable model. Estimating generation costs from observed schedules is then formulated as Bayesian inference. A prior distribution encodes an initial belief on parameters, and the inference consists of updating the belief with the posterior distribution given observations. We use balanced neural posterior estimation (BNPE) to learn this posterior. Validation on the IEEE RTS-96 test system shows that marginal costs are recovered with narrow credible intervals, while start-up costs remain largely unidentifiable from schedules alone. The method is benchmarked against an inverse-optimization algorithm that exhibits larger parameter errors without uncertainty quantification.

eess.SY↗

arXiv2025

Enforcing governing equation constraints in neural PDE solvers via training-free projections

Neural PDE solvers used for scientific simulation often violate governing equation constraints. While linear constraints can be projected cheaply, many constraints are nonlinear, complicating projection onto the feasible set. Dynamical PDEs are especially difficult because constraints induce long-range dependencies in time. In this work, we evaluate two training-free, post hoc projections of approximate solutions: a nonlinear optimization-based projection, and a local linearization-based projection using Jacobian-vector and vector-Jacobian products. We analyze constraints across representative PDEs and find that both projections substantially reduce violations and improve accuracy over physics-informed baselines.

cs.LG↗

arXiv2025

Appa: Bending Weather Dynamics with Latent Diffusion Models for Global Data Assimilation

Deep learning has advanced weather forecasting, but accurate predictions first require identifying the current state of the atmosphere from observational data. In this work, we introduce Appa, a score-based data assimilation model generating global atmospheric trajectories at 0.25\si{\degree} resolution and 1-hour intervals. Powered by a 565M-parameter latent diffusion model trained on ERA5, Appa can be conditioned on arbitrary observations to infer plausible trajectories, without retraining. Our probabilistic framework handles reanalysis, filtering, and forecasting, within a single model, producing physically consistent reconstructions from various inputs. Results establish latent score-based data assimilation as a promising foundation for future global atmospheric modeling systems.

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

Training-Free Data Assimilation with GenCast

Data assimilation is widely used in many disciplines such as meteorology, oceanography, and robotics to estimate the state of a dynamical system from noisy observations. In this work, we propose a lightweight and general method to perform data assimilation using diffusion models pre-trained for emulating dynamical systems. Our method builds on particle filters, a class of data assimilation algorithms, and does not require any further training. As a guiding example throughout this work, we illustrate our methodology on GenCast, a diffusion-based model that generates global ensemble weather forecasts.

cs.LG↗

arXiv2025

Turbulent Injection assisted by Diffusion Models for Scale Resolving Simulations

The present research proposes a new memory-efficient method using diffusion models to inject turbulent inflow conditions into Large Eddy Simulation (LES) and Direct Numerical Simulation (DNS) for various flow problems. A guided diffusion model was trained on Decaying Homogeneous Isotropic Turbulence (DHIT) samples, characterized by different turbulent kinetic energy levels and integral length scales. Samples generated by the diffusion model accurately reproduce turbulence statistics, such as the energy spectrum and the two-point autocorrelation functions, while preserving the ability to generate instantaneous three-dimensional velocity fields with detailed fluctuations. Physical representativeness is also evaluated by injecting the \textit{synthetic} samples into a free domain (i.e., without any wall boundary) through an inlet boundary condition. The method demonstrates promising results regarding energy spectrum, spatial correlation, turbulent kinetic energy level, and integral length scales without increasing the development distance as compared to a library-based method. The following article has been submitted to/accepted by Physics of Fluid (AIP Publishing LLC). After publication, it will be available https://doi.org/10.1063/5.0278541.

physics.flu-dyn↗

arXiv2025

Panchromatic characterization of the Y0 brown dwarf WISEP J173835.52+273258.9 using JWST/MIRI

Cold brown dwarf atmospheres are good training grounds for analyzing temperate giant planets. WISEP J173835.52+273258.9 (WISE 1738) is an isolated Y0 brown dwarf with a temperature between 350-400 K, at the T-Y transition. While its near-infrared spectrum has been studied, bulk properties and chemistry remain uncertain. We analyze new JWST MIRI medium-resolution spectra (5-18 micron), combined with near-infrared spectra (0.98-2.2 micron) from HST/WFC3 and Gemini/GNIRS, to better constrain WISE 1738's atmosphere and physical parameters. We use Neural Posterior Estimation (NPE) with a cloud-free petitRADTRANS model and evaluate results using posterior checks, coverage, and L-C2ST diagnostics. Our retrieval confirms previous constraints on H2O, CH4, and NH3, and for the first time constrains CO, CO2, and 15NH3. We find evidence of disequilibrium chemistry through CO and CO2 abundances not expected under equilibrium. Estimated properties are temperature 402 (+12,-9) K, log g 4.43 (+0.26,-0.34) cm/s2, mass 13 (+11,-7) M_Jup, radius 1.14 (+0.03,-0.03) R_Jup, and bolometric luminosity -6.52 (+0.05,-0.04) log L/L_sun. Evolutionary models suggest an age between 1 and 4 Gyr, consistent with a 6-hour rotation. We place an upper bound on 15NH3, implying a 3-sigma lower limit on the 14N/15N ratio of 275. We also derive a C/O ratio of 1.35 (+0.39,-0.31) and metallicity of 0.34 (+0.12,-0.11), without accounting for oxygen sequestration.

astro-ph.EP↗

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

Video-Driven Graph Network-Based Simulators

Lifelike visualizations in design, cinematography, and gaming rely on precise physics simulations, typically requiring extensive computational resources and detailed physical input. This paper presents a method that can infer a system's physical properties from a short video, eliminating the need for explicit parameter input, provided it is close to the training condition. The learned representation is then used within a Graph Network-based Simulator to emulate the trajectories of physical systems. We demonstrate that the video-derived encodings effectively capture the physical properties of the system and showcase a linear dependence between some of the encodings and the system's motion.

cs.CV↗

arXiv2025

A Neural Material Point Method for Particle-based Emulation

Mesh-free Lagrangian methods are widely used for simulating fluids, solids, and their complex interactions due to their ability to handle large deformations and topological changes. These physics simulators, however, require substantial computational resources for accurate simulations. To address these issues, deep learning emulators promise faster and scalable simulations, yet they often remain expensive and difficult to train, limiting their practical use. Inspired by the Material Point Method (MPM), we present NeuralMPM, a neural emulation framework for particle-based simulations. NeuralMPM interpolates Lagrangian particles onto a fixed-size grid, computes updates on grid nodes using image-to-image neural networks, and interpolates back to the particles. Similarly to MPM, NeuralMPM benefits from the regular voxelized representation to simplify the computation of the state dynamics, while avoiding the drawbacks of mesh-based Eulerian methods. We demonstrate the advantages of NeuralMPM on several datasets, including fluid dynamics and fluid-solid interactions. Compared to existing methods, NeuralMPM reduces training times from days to hours, while achieving comparable or superior long-term accuracy, making it a promising approach for practical forward and inverse problems. A project page is available at https://neuralmpm.isach.be

cs.LG↗

arXiv2024

Cost Estimation in Unit Commitment Problems Using Simulation-Based Inference

The Unit Commitment (UC) problem is a key optimization task in power systems to forecast the generation schedules of power units over a finite time period by minimizing costs while meeting demand and technical constraints. However, many parameters required by the UC problem are unknown, such as the costs. In this work, we estimate these unknown costs using simulation-based inference on an illustrative UC problem, which provides an approximated posterior distribution of the parameters given observed generation schedules and demands. Our results highlight that the learned posterior distribution effectively captures the underlying distribution of the data, providing a range of possible values for the unknown parameters given a past observation. This posterior allows for the estimation of past costs using observed past generation schedules, enabling operators to better forecast future costs and make more robust generation scheduling forecasts. We present avenues for future research to address overconfidence in posterior estimation, enhance the scalability of the methodology and apply it to more complex UC problems modeling the network constraints and renewable energy sources.

cs.LG↗