259 citations · 978 across the 60 of their papers we have counts for
66 papers
Context Copying Modulation: The Role of Entropy Neurons in Managing Parametric and Contextual Knowledge Conflicts
Zineddine Tighidet, Andrea Mogini, Hedi Ben-younes +3
The behavior of Large Language Models (LLMs) when facing contextual information that conflicts with their internal parametric knowledge is inconsistent, with no generally accepted…
Probing Language Models on Their Knowledge Source
Zineddine Tighidet, Andrea Mogini, Jiali Mei +2
Large Language Models (LLMs) often encounter conflicts between their learned, internal (parametric knowledge, PK) and external knowledge provided during inference (contextual knowl…
MEXMA: Token-level objectives improve sentence representations
João Maria Janeiro, Benjamin Piwowarski, Patrick Gallinari +1
Current pre-trained cross-lingual sentence encoders approaches use sentence-level objectives only. This can lead to loss of information, especially for tokens, which then degrades…
AROMA: Preserving Spatial Structure for Latent PDE Modeling with Local Neural Fields
Louis Serrano, Thomas X Wang, Etienne Le Naour +2
We present AROMA (Attentive Reduced Order Model with Attention), a framework designed to enhance the modeling of partial differential equations (PDEs) using local neural fields. Ou…
Module-wise Training of Neural Networks via the Minimizing Movement Scheme
Skander Karkar, Ibrahim Ayed, Emmanuel de Bézenac +1
Greedy layer-wise or module-wise training of neural networks is compelling in constrained and on-device settings where memory is limited, as it circumvents a number of problems of…
INFINITY: Neural Field Modeling for Reynolds-Averaged Navier-Stokes Equations
Louis Serrano, Leon Migus, Yuan Yin +2
For numerical design, the development of efficient and accurate surrogate models is paramount. They allow us to approximate complex physical phenomena, thereby reducing the computa…
Adversarial Sample Detection Through Neural Network Transport Dynamics
Skander Karkar, Patrick Gallinari, Alain Rakotomamonjy
We propose a detector of adversarial samples that is based on the view of neural networks as discrete dynamic systems. The detector tells clean inputs from abnormal ones by compari…
Stability of implicit neural networks for long-term forecasting in dynamical systems
Leon Migus, Julien Salomon, Patrick Gallinari
Forecasting physical signals in long time range is among the most challenging tasks in Partial Differential Equations (PDEs) research. To circumvent limitations of traditional solv…
Operator Learning with Neural Fields: Tackling PDEs on General Geometries
Louis Serrano, Lise Le Boudec, Armand Kassaï Koupaï +4
Machine learning approaches for solving partial differential equations require learning mappings between function spaces. While convolutional or graph neural networks are constrain…
Time Series Continuous Modeling for Imputation and Forecasting with Implicit Neural Representations
Etienne Le Naour, Louis Serrano, Léon Migus +5
We introduce a novel modeling approach for time series imputation and forecasting, tailored to address the challenges often encountered in real-world data, such as irregular sample…
Learning from Multiple Sources for Data-to-Text and Text-to-Data
Song Duong, Alberto Lumbreras, Mike Gartrell +1
Data-to-text (D2T) and text-to-data (T2D) are dual tasks that convert structured data, such as graphs or tables into fluent text, and vice versa. These tasks are usually handled se…
Exemplars and Counterexemplars Explanations for Image Classifiers, Targeting Skin Lesion Labeling
Carlo Metta, Riccardo Guidotti, Yuan Yin +2
Explainable AI consists in developing mechanisms allowing for an interaction between decision systems and humans by making the decisions of the formers understandable. This is part…
AirfRANS: High Fidelity Computational Fluid Dynamics Dataset for Approximating Reynolds-Averaged Navier-Stokes Solutions
Florent Bonnet, Ahmed Jocelyn Mazari, Paola Cinnella +1
Surrogate models are necessary to optimize meaningful quantities in physical dynamics as their recursive numerical resolutions are often prohibitively expensive. It is mainly the c…
HMOE: Hypernetwork-based Mixture of Experts for Domain Generalization
Jingang Qu, Thibault Faney, Ze Wang +3
Due to domain shifts, machine learning systems typically struggle to generalize well to new domains that differ from those of training data, which is what domain generalization (DG…
Block-wise Training of Residual Networks via the Minimizing Movement Scheme
Skander Karkar, Ibrahim Ayed, Emmanuel de Bézenac +1
End-to-end backpropagation has a few shortcomings: it requires loading the entire model during training, which can be impossible in constrained settings, and suffers from three loc…
Continuous PDE Dynamics Forecasting with Implicit Neural Representations
Yuan Yin, Matthieu Kirchmeyer, Jean-Yves Franceschi +2
Effective data-driven PDE forecasting methods often rely on fixed spatial and / or temporal discretizations. This raises limitations in real-world applications like weather predict…
An extensible Benchmarking Graph-Mesh dataset for studying Steady-State Incompressible Navier-Stokes Equations
Florent Bonnet, Jocelyn Ahmed Mazari, Thibaut Munzer +2
Recent progress in \emph{Geometric Deep Learning} (GDL) has shown its potential to provide powerful data-driven models. This gives momentum to explore new methods for learning phys…
Multi-scale Physical Representations for Approximating PDE Solutions with Graph Neural Operators
Léon Migus, Yuan Yin, Jocelyn Ahmed Mazari +1
Representing physical signals at different scales is among the most challenging problems in engineering. Several multi-scale modeling tools have been developed to describe physical…
PTFlash : A deep learning framework for isothermal two-phase equilibrium calculations
Jingang Qu, Thibault Faney, Jean-Charles de Hemptinne +2
Phase equilibrium calculations are an essential part of numerical simulations of multi-component multi-phase flow in porous media, accounting for the largest share of the computati…
Diverse Weight Averaging for Out-of-Distribution Generalization
Alexandre Ramé, Matthieu Kirchmeyer, Thibaud Rahier +3
Standard neural networks struggle to generalize under distribution shifts in computer vision. Fortunately, combining multiple networks can consistently improve out-of-distribution…
Generalizing to New Physical Systems via Context-Informed Dynamics Model
Matthieu Kirchmeyer, Yuan Yin, Jérémie Donà +3
Data-driven approaches to modeling physical systems fail to generalize to unseen systems that share the same general dynamics with the learning domain, but correspond to different…
Explainable Deep Image Classifiers for Skin Lesion Diagnosis
Carlo Metta, Andrea Beretta, Riccardo Guidotti +4
A key issue in critical contexts such as medical diagnosis is the interpretability of the deep learning models adopted in decision-making systems. Research in eXplainable Artificia…
Mapping conditional distributions for domain adaptation under generalized target shift
Matthieu Kirchmeyer, Alain Rakotomamonjy, Emmanuel de Bezenac +1
We consider the problem of unsupervised domain adaptation (UDA) between a source and a target domain under conditional and label shift a.k.a Generalized Target Shift (GeTarS). Unli…
Separating Retention from Extraction in the Evaluation of End-to-end Relation Extraction
Bruno Taillé, Vincent Guigue, Geoffrey Scoutheeten +1
State-of-the-art NLP models can adopt shallow heuristics that limit their generalization capability (McCoy et al., 2019). Such heuristics include lexical overlap with the training…