paper-conference

Mitigating Text Toxicity with Counterfactual Generation
Toxicity mitigation consists in rephrasing text in order to remove offensive or harmful meaning. Neural natural language processing …
Milan Bhan, Jean-Noël Vittaut, Nina Achache, Victor Legrand, Annabelle Blangero, Nicolas Chesneau, Juliette Murris, Marie-Jeanne Lesot
Citation
Towards Achieving Concept Completeness for Textual Concept Bottleneck Models
This paper proposes Complete Textual Concept Bottleneck Model (CT-CBM), a novel TCBM generator building concept labels in a fully …
Milan Bhan, Yann Choho, Jean-Noël Vittaut, Nicolas Chesneau, Pierre Moreau, Marie-Jeanne Lesot
Citation DOI URL
Assessing the Interpretability of Fuzzy Rule Bases
Fuzzy systems are valued for their ability to model complex phenomena while contributing to maintaining inter-pretability through …
Thomas Pontoizeau, Romain Caillière, Marie-Jeanne Lesot, Christophe Marsala, Nicolas Museux, Jean-Noël Vittaut
Citation DOI
Self-AMPLIFY: Improving Small Language Models with Self Post Hoc Explanations
Incorporating natural language rationales in the prompt and In-Context Learning (ICL) have led to a significant improvement of Large …
Milan Bhan, Jean-Noël Vittaut, Nicolas Chesneau, Marie-Jeanne Lesot
Citation DOI URL
AROMA: Preserving Spatial Structure for Latent PDE Modeling with Local Neural Fields
We present AROMA (Attentive Reduced Order Model with Attention), a framework designed to enhance the modeling of partial differential …
Louis Serrano, Thomas X Wang, Etienne Le Naour, Jean-Noël Vittaut, Patrick Gallinari
PDF Citation
Boosting Generalization in Parametric PDE Neural Solvers through Adaptive Conditioning
Solving parametric partial differential equations (PDEs) presents significant challenges for data-driven methods due to the sensitivity …
Armand Kassaï Koupaï, Jorge Mifsut Benet, Yuan Yin, Jean-Noël Vittaut, Patrick Gallinari
PDF Citation
Latent Diffusion Transformer with Local Neural Field as PDE Surrogate Model
We introduce a diffusion transformer architecture, AROMA (Attentive Reduced Order Model with Attention), for dynamics modeling of …
Louis Serrano, Jean-Noël Vittaut, Patrick Gallinari
Citation URL
Zebra: a continuous generative transformer for solving parametric PDEs
Foundation models have revolutionized deep learning, moving beyond task-specific architectures to versatile models pre-trained using …
Louis Serrano, Pierre Erbacherand Jean-Noël Vittaut, Patrick Gallinari
Citation URL
Enhancing textual counterfactual explanation intelligibility through Counterfactual Feature Importance
Textual counterfactual examples explain a prediction by modifying the tokens of an initial instance in order to flip the outcome of a …
Milan Bhan, Jean-Noël Vittaut, Nicolas Chesneau, Marie-Jeanne Lesot
Citation DOI URL
Operator Learning on Free-Form Geometries
Operator Learning models usually rely on a fixed sampling scheme for training which might limit their ability to generalize to new …
Louis Serrano, Jean-Noël Vittaut, Patrick Gallinari
Citation URL