Gilles Louppe

Flow cytometry produces high-dimensional single-cell protein measurements central to immunophenotyping and clinical monitoring. Yet analysis still relies largely on manual gating, which is labour-intensive, poorly reproducible, and ill-suited to large marker panels. Existing computational approaches address classification or discovery in isolation, treating cell-type identity as a post-hoc annotation rather than as part of the generative model itself. We present MARVIN, a semi-supervised variational autoencoder that encodes the assumption that cells organise into discrete populations with continuous intra-population variability through a Gaussian mixture prior in the latent space. Because each component represents a distinct cell population, classification, discovery, and density estimation emerge as complementary views of the same representation. On public benchmarks, MARVIN matches or exceeds existing methods using as few as 10% labelled cells. Trained exclusively on healthy samples, it identifies leukaemic cells through elevated reconstruction error, providing an unsupervised anomaly detection signal. On paired stimulation data, it maintains stable population assignments while capturing condition-specific shifts in abundance and marker expression at patient-level resolution. MARVIN is open-source and designed for local deployment, adapting to institution-specific panels and instruments

Regional Climate Models (RCMs) provide high-resolution, physics-based fields, but they face three main limitations. First, they are computationally expensive and hence difficult to scale across scenarios or ensembles. Second, they lack uncertainty quantification. Third, they usually take only coarse data from Earth System Models (ESMs) or reanalysis to predict fields, without assimilating real observations. In response to these problems, neural emulators of RCMs have been developed over different regions. In this work, we present MAR.ia, a diffusion-based emulator of MAR, an RCM developed at ULiège tailored to Belgium (Doutreloup et al., 2019). Our approach maps coarse atmospheric and surface reanalysis variables (ERA5 at 0.25° and 1° resolution) to key surface variables (temperature, precipitation and wind speed) at the resolution of MAR (5 km). The emulator is conditioned on ERA5 reanalysis every six hours (as the forcing of MAR) in order to give hourly MAR-like fields. We assess the sensitivity of the emulator to the choice of ERA5 fields, identifying the key drivers to reproduce MAR dynamics. We solve the three main limitations initially stated: we reduce computational costs by several orders of magnitude, we estimate uncertainty by sampling several times for the same coarse inputs (generation of ensembles), and we incorporate observational constraints from ground stations and satellites directly during sampling, while showing competitive metrics, i.e. correlation of ~0.99 for the temperature at 2m. Future work will attempt to use ESM outputs (weather forecast or CMIP future projections) as context variables instead of reanalysis, enabling both short-term meteorological predictions and long-term climate projections up to 2100, over Belgium. Doutreloup, S., Wyard, C., Amory, C., Kittel, C., Erpicum, M., and Fettweis, X. (2019). Sensitivity to Convective Schemes on Precipitation Simulated by the Regional Climate Model MAR over Belgium (1987–2017), Atmosphere, 10( 1), 34. https://doi.org/10.3390/atmos10010034.

MAR is a Regional Climate Model (RCM) used over Belgium that provides deterministic downscaling of reanalyses and Earth System Models (ESMs) at 5-km resolution (Doutreloup et al., 2019). These high-resolution fields are computationally expensive to produce as they require solving complex physical equations. Combined with its deterministic nature, this limits the use of MAR for assessing the frequency and intensity of extreme events and their future changes.To address this limitation, we have developed MAR.ia, a diffusion-based emulator of MAR which provides probabilistic estimates of downscaled fields at a lower computational cost (from 0.25° and 1° ERA5 fields). This allows the direct generation of ensembles from which we can derive a range of possible weather outcomes and estimate their corresponding likelihood.However, the reproduction of extreme events is expected to be more challenging for diffusion models because these events might be scattered or absent from the training set. This is due to the fact that they are rare, and also to climate change which induces a shift between the training and testing distributions.We evaluate the MAR.ia reconstruction of extreme heatwaves, storms and heavy rainfall associated with several daily historical events in Belgium and compare these results with those obtained on average over the testing period.This evaluation enables us to critically assess the ability of deep generative models, and more precisely diffusion models approaches, to faithfully reconstruct out-of-distribution events. Doutreloup, S., Wyard, C., Amory, C., Kittel, C., Erpicum, M., and Fettweis, X. (2019). Sensitivity to Convective Schemes on Precipitation Simulated by the Regional Climate Model MAR over Belgium (1987–2017), Atmosphere, 10, 34. https://doi.org/10.3390/atmos10010034.

Context . Cold brown dwarf atmospheres provide a good training ground for the analysis of atmospheres of temperate giant planets. WISEP J173835.52+273258.9 (WISE 1738) is an isolated cold brown dwarf and a Y0 spectral standard with a temperature between 350-400 K, lying at the boundary of the T-Y transition. Although its atmosphere has been extensively studied in the near-infrared, its bulk physical parameters and atmospheric chemistry and dynamics are not well understood. Aims . Using a Mid-Infrared Instrument (MIRI) medium-resolution spectrum (5-18 μm), combined with near-infrared spectra (0.982.2 μm) from Hubble Space Telescope’s (HST) Wide Field Camera 3 (WFC3) and Gemini Observatory’s Near-Infrared Spectrograph (GNIRS), we aim to accurately characterize the atmospheric chemistry and bulk physical parameters of WISE 1738. Methods . We perform a combined atmospheric retrieval on the MIRI, GNIRS, and WFC3 spectra using a machine learning algorithm called Neural Posterior Estimation (NPE) assuming a cloud-free model implemented using petitRADTRANS . We demonstrate how this combined retrieval approach ensures robust constraints on the abundances of major atmospheric species, the pressure-temperature ( P - T ) profile, bulk C/O, and metallicity [M/H], along with bulk physical properties such as effective temperature, radius, surface gravity, mass, and luminosity. We estimate 1D and 2D marginal posterior distributions for the constrained parameters and evaluate our results using several qualitative and quantitative Bayesian diagnostics, including Local Classifier 2-Sample Test (L-C2ST), coverage, and posterior predictive checks. Results . The combined atmospheric retrieval confirms previous constraints on H ₂ O, CH ₄ , NH ₃ , and for the first time provides constraints on CO, CO ₂ , and ¹⁵ NH ₃ . It also gives better constraints on the physical parameters and the P - T profile while also revealing potential biases in characterizing objects using data from limited wavelength ranges. The retrievals further suggest the presence of disequilibrium chemistry, as evidenced by the constrained abundances of CO and CO ₂ , which are otherwise expected to be depleted and hence not visible beyond the near-infrared wavelengths under equilibrium conditions. We estimate the physical parameters of the object as follows: an effective temperature of 402 ₋₉ ⁺¹² K, surface gravity (log g ) of 4.43 _−0.34 ^+0.26 cm s ⁻² , mass of 13 ₋₇ ⁺¹¹ M _Jup , radius of 1.14 _−0.03 ^+0.03 R _Jup , and a bolometric luminosity of −6.52 _−0.04 ^+0.05 log L/L _⊙ . Based on these values, the evolutionary models suggest an age between 1 and 4 Gyr, which is consistent with a high rotation rate of 6 h of the brown dwarf. We further obtain an upper bound on the ¹⁵ NH ₃ abundance, enabling a 3σ lower bound calculation of the ¹⁴ N/ ¹⁵ N ratio = 275, unable to interpret the formation pathway as core collapse. Additionally, we calculate a C/O ratio of 1.35 _−0.31 ^+0.39 and a metallicity of 0.34 _−0.11 ^+0.12 without considering any oxygen sequestration effects.

The present research proposes a new memory-efficient method using diffusion models to inject turbulent inflow conditions into large eddy simulation and direct numerical simulation for various flow problems. A guided diffusion model was trained on decaying homogeneous isotropic turbulence 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 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.

Context. Standard cosmological analysis, which is based on two-point statistics, fails to extract all the information embedded in the cosmological data. This limits our ability to precisely constrain cosmological parameters. Through willingness to use modern analysis techniques to match the power of upcoming telescopes, 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. Aims. We focused on full-field inference, which is considered the optimal form of inference as it enables the recovery of cosmological constraints from simulations without any loss of cosmological information. Our objective is to review and benchmark several ways of conducting full-field inference to gain insight into the number of simulations required for each method. Specifically, we made a distinction between explicit inference methods that require an explicit form of the likelihood, such that it can be evaluated and thus sampled through sampling schemes and implicit inference methods that can be used when only an implicit version of the likelihood is available through simulations. Moreover, it is crucial for explicit full-field inference to use a differentiable forward model. Similarly, we aim to discuss the advantages of having differentiable forward models for implicit full-field inference. Methods. We used the sbi_lens package (https://github.com/DifferentiableUniverseInitiative/sbi_lens), which provides a fast and differentiable log-normal forward model to generate convergence maps mimicking a simplified version of LSST Y10 quality. While the analyses use a simplified forward model, the goal is to illustrate key methodologies and their implications. Specifically, this fast-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 (NUTS) sampler. The latter starts by compressing the data into sufficient statistics and uses the neural likelihood estimation (NLE) algorithm and the one augmented with gradient (∂NLE) to learn the likelihood distribution and then sample the posterior distribution. Results. We performed a full-field analysis on LSST Y10-like weak-lensing-simulated log-normal convergence maps, where we constrain (Ω_c,Ω_b,σ₈,h₀,n_s,w₀). We demonstrate that explicit full-field and implicit full-field inference yield consistent constraints. Explicit full-field inference requires 630 000 simulations with our particular sampler, which corresponds to 400 independent samples. Implicit full-field inference requires a maximum of 101 000 simulations split into 100 000 simulations to build neural-based sufficient statistics (this number of simulations is not fine-tuned) and 1000 simulations to perform inference using implicit inference. Additionally, while differentiability is very useful for explicit full-field inference, we show that, for this specific case, our way of exploiting the gradients does not help implicit full-field inference significantly.

We introduce a novel approach to 6-DoF robotic grasping based on simulation-based inference. Our approach combines sequential neural ratio estimation with a neural implicit representation for the Bayesian inference of hand configurations in cluttered environments. We propose to compute the maximum a posteriori by gradient descent, more specifically using Riemannian gradient descent, to preserve the geometry of the rotation space and capitalize on the full differentiability of our model. We demonstrate the capabilities of our approach on a grasping benchmark both in simulation and on a real robot. Our performance generalizes well across different scenarios, achieving high success rates.

To understand high precision observations of exoplanets and brown dwarfs, we need detailed and complex general circulation models (GCMs) that incorporate hydrodynamics, chemistry, and radiation. For this study, we specifically examined the coupling between chemistry and radiation in GCMs and compared different methods for the mixing of opacities of different chemical species in the correlated-k assumption, when equilibrium chemistry cannot be assumed. We propose a fast machine learning method based on DeepSets (DS), which effectively combines individual correlated-k opacities (k-tables). We evaluated the DS method alongside other published methods such as adaptive equivalent extinction (AEE) and random overlap with rebinning and resorting (RORR). We integrated these mixing methods into our GCM (expeRT/MITgcm) and assessed their accuracy and performance for the example of the hot Jupiter HD 209458 b. Our findings indicate that the DS method is both accurate and efficient for GCM usage, whereas RORR is too slow. Additionally, we observed that the accuracy of AEE depends on its specific implementation and may introduce numerical issues in achieving radiative transfer solution convergence. We then applied the DS mixing method in a simplified chemical disequilibrium situation, where we modeled the rainout of TiO and VO, and confirmed that the rainout of TiO and VO would hinder the formation of a stratosphere. To further expedite the development of consistent disequilibrium chemistry calculations in GCMs, we provide documentation and code for coupling the DS mixing method with correlated-k radiative transfer solvers. The DS method has been extensively tested to be accurate enough for GCMs; however, other methods might be needed for accelerating atmospheric retrievals.