no code implementations • 4 Mar 2024 • Marion Ullmo, Nabila Aghnim, Aurélien Decelle, Miguel Aragon-Calvo
Cosmological simulations play a key role in the prediction and understanding of large scale structure formation from initial conditions.
1 code implementation • 5 Sep 2023 • Aurélien Decelle, Cyril Furtlehner, Alfonso De Jesus Navas Gómez, Beatriz Seoane
Generative models offer a direct way of modeling complex data.
no code implementations • 7 Aug 2023 • Giovanni Catania, Aurélien Decelle, Beatriz Seoane
We characterize the equilibrium properties of a model of $y$ coupled binary perceptrons in the teacher-student scenario, subject to a learning rule, with an explicit ferromagnetic coupling proportional to the Hamming distance between the students' weights.
1 code implementation • 13 Jul 2023 • Alessandra Carbone, Aurélien Decelle, Lorenzo Rosset, Beatriz Seoane
In this study, we address the challenge of using energy-based models to produce high-quality, label-specific data in complex structured datasets, such as population genetics, RNA or protein sequences data.
1 code implementation • 3 Feb 2023 • Aurélien Decelle, Lorenzo Rosset, Beatriz Seoane
Datasets in the real world are often complex and to some degree hierarchical, with groups and sub-groups of data sharing common characteristics at different levels of abstraction.
no code implementations • 23 Jan 2023 • Elisabeth Agoritsas, Giovanni Catania, Aurélien Decelle, Beatriz Seoane
In this paper, we quantify the impact of using non-convergent Markov chains to train Energy-Based models (EBMs).
no code implementations • 17 Nov 2022 • Adriano Barra, Giovanni Catania, Aurélien Decelle, Beatriz Seoane
Introduced by Kosko in 1988 as a generalization of the Hopfield model to a bipartite structure, the simplest architecture is defined by two layers of neurons, with synaptic connections only between units of different layers: even without internal connections within each layer, information storage and retrieval are still possible through the reverberation of neural activities passing from one layer to another.
1 code implementation • 2 Jun 2022 • Nicolas Béreux, Aurélien Decelle, Cyril Furtlehner, Beatriz Seoane
Moreover, we show that this sampling technique can also be used to improve the computation of the log-likelihood gradient during training, leading to dramatic improvements in training RBMs with artificial clustered datasets.
1 code implementation • 16 Jun 2021 • Tony Bonnaire, Aurélien Decelle, Nabila Aghanim
A regularized version of Mixture Models is proposed to learn a principal graph from a distribution of $D$-dimensional data points.
1 code implementation • NeurIPS 2021 • Aurélien Decelle, Cyril Furtlehner, Beatriz Seoane
In this work, we show that this mixing time plays a crucial role in the dynamics and stability of the trained model, and that RBMs operate in two well-defined regimes, namely equilibrium and out-of-equilibrium, depending on the interplay between this mixing time of the model and the number of steps, $k$, used to approximate the gradient.
no code implementations • 23 Nov 2020 • Aurélien Decelle, Cyril Furtlehner
This review deals with Restricted Boltzmann Machine (RBM) under the light of statistical physics.
no code implementations • 10 Nov 2020 • Marion Ullmo, Aurélien Decelle, Nabila Aghanim
We find that the GAN successfully generates new images that are statistically consistent with the images it was trained on.
no code implementations • 19 Dec 2019 • Giancarlo Fissore, Aurélien Decelle, Cyril Furtlehner, Yufei Han
In order to take full advantage of these dependencies we consider a purely probabilistic setting in which the features imputation and multi-label classification problems are jointly solved.
no code implementations • 31 Oct 2019 • Aurélien Decelle, Cyril Furtlehner
We consider a special type of Restricted Boltzmann machine (RBM), namely a Gaussian-spherical RBM where the visible units have Gaussian priors while the vector of hidden variables is constrained to stay on an ${\mathbbm L}_2$ sphere.
1 code implementation • 16 Apr 2019 • Aurélien Decelle, Victor Martin-Mayor, Beatriz Seoane
We explore the capacity of neural networks to detect a symmetry with complex local and non-local patterns : the gauge symmetry Z 2 .
no code implementations • 5 Mar 2018 • Aurélien Decelle, Giancarlo Fissore, Cyril Furtlehner
In the non-linear regime, instead, the selected modes interact and eventually impose a matching of the order parameters to their empirical counterparts estimated from the data.
no code implementations • 9 Aug 2017 • Aurélien Decelle, Giancarlo Fissore, Cyril Furtlehner
This let us show how in the linear regime, in which the RBM is found to operate at the beginning of the training, the statistical properties of the data drive the selection of the unstable modes of the weight matrix.