Pierre Ablin

90 papers A* 22A 5B 1Misc 2Journal 57Unranked 3
YearRankTypeTitle / Venue / Authors
2026 J jnl
LaCy: What Small Language Models Can and Should Learn is Not Just a Question of Loss.
CoRR
Szilvia Ujváry, Louis Béthune, Pierre Ablin, João Monteiro, Marco Cuturi, Michael Kirchhof
2026 J jnl
MVICAD
IEEE Trans. Biomed. Eng.
Ambroise Heurtebise, Omar Chehab, Pierre Ablin, Alexandre Gramfort
2026 J jnl
Optimal Splitting of Language Models from Mixtures to Specialized Domains.
CoRR
Skyler Seto, Pierre Ablin, Anastasiia Filippova, Jiayuan Ye, Louis Béthune, Angelos Katharopoulos, David Grangier
2026 J jnl
The Design Space of Tri-Modal Masked Diffusion Models.
CoRR
Louis Béthune, Victor Turrisi, Bruno Kacper Mlodozeniec, Pau Rodríguez López, Lokesh Boominathan, Nikhil Bhendawade, Amitis Shidani, Joris Pelemans, Theo X. Olausson, R. Devon Hjelm, Paul Dixon, João Monteiro, Pierre Ablin, Vishnu Banna, Arno Blaas, Nick Henderson, Kari Noriy, Dan Busbridge, Joshua M. Susskind, Marco Cuturi, Irina Belousova, Luca Zappella, Russ Webb, Jason Ramapuram
2025 J jnl
A Unified Perspective on the Dynamics of Deep Transformers.
CoRR
Valérie Castin, Pierre Ablin, José Antonio Carrillo, Gabriel Peyré
2025 J jnl
Completed Hyperparameter Transfer across Modules, Width, Depth, Batch and Duration.
CoRR
Bruno Mlodozeniec, Pierre Ablin, Louis Béthune, Dan Busbridge, Michal Klein, Jason Ramapuram, Marco Cuturi
2025 J jnl
Identifiable Multi-View Causal Discovery Without Non-Gaussianity.
CoRR
Ambroise Heurtebise, Omar Chehab, Pierre Ablin, Alexandre Gramfort, Aapo Hyvärinen
2025 J jnl
Learning Unmasking Policies for Diffusion Language Models.
CoRR
Metod Jazbec, Theo X. Olausson, Louis Béthune, Pierre Ablin, Michael Kirchhof, João Monteiro, Victor Turrisi, Jason Ramapuram, Marco Cuturi
2025 J jnl
MVICAD2: Multi-View Independent Component Analysis with Delays and Dilations.
CoRR
Ambroise Heurtebise, Omar Chehab, Pierre Ablin, Alexandre Gramfort
2025 A* conf
Scaling Laws for Forgetting during Finetuning with Pretraining Data Injection.
ICML
Louis Béthune, David Grangier, Dan Busbridge, Eleonora Gualdoni, Marco Cuturi, Pierre Ablin
2025 J jnl
Scaling Laws for Forgetting during Finetuning with Pretraining Data Injection.
CoRR
Louis Béthune, David Grangier, Dan Busbridge, Eleonora Gualdoni, Marco Cuturi, Pierre Ablin
2025 J jnl
Scaling Laws for Optimal Data Mixtures.
CoRR
Mustafa Shukor, Louis Béthune, Dan Busbridge, David Grangier, Enrico Fini, Alaaeldin El-Nouby, Pierre Ablin
2025 A* conf
Shielded Diffusion: Generating Novel and Diverse Images using Sparse Repellency.
ICML
Michael Kirchhof, James Thornton, Louis Béthune, Pierre Ablin, Eugène Ndiaye, Marco Cuturi
2025 A* conf
Soup-of-Experts: Pretraining Specialist Models via Parameters Averaging.
ICML
Pierre Ablin, Angelos Katharopoulos, Skyler Seto, David Grangier
2025 J jnl
Soup-of-Experts: Pretraining Specialist Models via Parameters Averaging.
CoRR
Pierre Ablin, Angelos Katharopoulos, Skyler Seto, David Grangier
2025 A* conf
Task-Adaptive Pretrained Language Models via Clustered-Importance Sampling.
ICLR
David Grangier, Simin Fan, Skyler Seto, Pierre Ablin
2025 A* conf
The AdEMAMix Optimizer: Better, Faster, Older.
ICLR
Matteo Pagliardini, Pierre Ablin, David Grangier
2025 J jnl
The Data-Quality Illusion: Rethinking Classifier-Based Quality Filtering for LLM Pretraining.
CoRR
Thiziri Nait Saada, Louis Béthune, Michal Klein, David Grangier, Marco Cuturi, Pierre Ablin
2025 J jnl
The Geometries of Truth Are Orthogonal Across Tasks.
CoRR
Waïss Azizian, Michael Kirchhof, Eugène Ndiaye, Louis Béthune, Michal Klein, Pierre Ablin, Marco Cuturi
2025 A* conf
Theory, Analysis, and Best Practices for Sigmoid Self-Attention.
ICLR
Jason Ramapuram, Federico Danieli, Eeshan Gunesh Dhekane, Floris Weers, Dan Busbridge, Pierre Ablin, Tatiana Likhomanenko, Jagrit Digani, Zijin Gu, Amitis Shidani, Russell Webb
2024 A conf
A Lower Bound and a Near-Optimal Algorithm for Bilevel Empirical Risk Minimization.
AISTATS
Mathieu Dagréou, Thomas Moreau, Samuel Vaiter, Pierre Ablin
2024 J jnl
Adaptive Training Distributions with Scalable Online Bilevel Optimization.
Trans. Mach. Learn. Res.
David Grangier, Pierre Ablin, Awni Hannun
2024 A* conf
Careful with that Scalpel: Improving Gradient Surgery with an EMA.
ICML
Yu-Guan Hsieh, James Thornton, Eugène Ndiaye, Michal Klein, Marco Cuturi, Pierre Ablin
2024 J jnl
Careful with that Scalpel: Improving Gradient Surgery with an EMA.
CoRR
Yu-Guan Hsieh, James Thornton, Eugène Ndiaye, Michal Klein, Marco Cuturi, Pierre Ablin
2024 J jnl
Dynamic Gradient Alignment for Online Data Mixing.
CoRR
Simin Fan, David Grangier, Pierre Ablin
2024 A conf
Enhancing Hypergradients Estimation: A Study of Preconditioning and Reparameterization.
AISTATS
Zhenzhang Ye, Gabriel Peyré, Daniel Cremers, Pierre Ablin
2024 J jnl
Enhancing Hypergradients Estimation: A Study of Preconditioning and Reparameterization.
CoRR
Zhenzhang Ye, Gabriel Peyré, Daniel Cremers, Pierre Ablin
2024 A* conf
How Smooth Is Attention?
ICML
Valérie Castin, Pierre Ablin, Gabriel Peyré
2024 J jnl
Infeasible Deterministic, Stochastic, and Variance-Reduction Algorithms for Optimization under Orthogonality Constraints.
J. Mach. Learn. Res.
Pierre Ablin, Simon Vary, Bin Gao, Pierre-Antoine Absil
2024 A* conf
Learning Elastic Costs to Shape Monge Displacements.
NeurIPS
Michal Klein, Aram-Alexandre Pooladian, Pierre Ablin, Eugène Ndiaye, Jonathan Niles-Weed, Marco Cuturi
2024 A* conf
Optimization without Retraction on the Random Generalized Stiefel Manifold.
ICML
Simon Vary, Pierre Ablin, Bin Gao, Pierre-Antoine Absil
2024 J jnl
Optimization without Retraction on the Random Generalized Stiefel Manifold.
CoRR
Simon Vary, Pierre Ablin, Bin Gao, Pierre-Antoine Absil
2024 J jnl
Sparse Repellency for Shielded Generation in Text-to-image Diffusion Models.
CoRR
Michael Kirchhof, James Thornton, Pierre Ablin, Louis Béthune, Eugène Ndiaye, Marco Cuturi
2024 J jnl
Specialized Language Models with Cheap Inference from Limited Domain Data.
CoRR
David Grangier, Angelos Katharopoulos, Pierre Ablin, Awni Hannun
2024 J jnl
Task-Adaptive Pretrained Language Models via Clustered-Importance Sampling.
CoRR
David Grangier, Simin Fan, Skyler Seto, Pierre Ablin
2024 J jnl
The AdEMAMix Optimizer: Better, Faster, Older.
CoRR
Matteo Pagliardini, Pierre Ablin, David Grangier
2024 J jnl
Theory, Analysis, and Best Practices for Sigmoid Self-Attention.
CoRR
Jason Ramapuram, Federico Danieli, Eeshan Gunesh Dhekane, Floris Weers, Dan Busbridge, Pierre Ablin, Tatiana Likhomanenko, Jagrit Digani, Zijin Gu, Amitis Shidani, Russ Webb
2023 J jnl
A Challenge in Reweighting Data with Bilevel Optimization.
CoRR
Anastasia Ivanova, Pierre Ablin
2023 J jnl
A Near-Optimal Algorithm for Bilevel Empirical Risk Minimization.
CoRR
Mathieu Dagréou, Thomas Moreau, Samuel Vaiter, Pierre Ablin
2023 J jnl
Adaptive Training Distributions with Scalable Online Bilevel Optimization.
CoRR
David Grangier, Pierre Ablin, Awni Hannun
2023 A* conf
How to Scale Your EMA.
NeurIPS
Dan Busbridge, Jason Ramapuram, Pierre Ablin, Tatiana Likhomanenko, Eeshan Gunesh Dhekane, Xavier Suau Cuadros, Russell Webb
2023 J jnl
How to Scale Your EMA.
CoRR
Dan Busbridge, Jason Ramapuram, Pierre Ablin, Tatiana Likhomanenko, Eeshan Gunesh Dhekane, Xavier Suau, Russ Webb
2023 J jnl
Infeasible Deterministic, Stochastic, and Variance-Reduction Algorithms for Optimization under Orthogonality Constraints.
CoRR
Pierre Ablin, Simon Vary, Bin Gao, Pierre-Antoine Absil
2023 J jnl
Learning Costs for Structured Monge Displacements.
CoRR
Michal Klein, Aram-Alexandre Pooladian, Pierre Ablin, Eugène Ndiaye, Jonathan Niles-Weed, Marco Cuturi
2023 A* conf
Monge, Bregman and Occam: Interpretable Optimal Transport in High-Dimensions with Feature-Sparse Maps.
ICML
Marco Cuturi, Michal Klein, Pierre Ablin
2023 J jnl
Monge, Bregman and Occam: Interpretable Optimal Transport in High-Dimensions with Feature-Sparse Maps.
CoRR
Marco Cuturi, Michal Klein, Pierre Ablin
2023 J jnl
MultiView Independent Component Analysis with Delays.
CoRR
Ambroise Heurtebise, Pierre Ablin, Alexandre Gramfort
2023 — conf
Multiview Independent Component Analysis with Delays.
MLSP
Ambroise Heurtebise, Pierre Ablin, Alexandre Gramfort
2023 J jnl
Test like you Train in Implicit Deep Learning.
CoRR
Zaccharie Ramzi, Pierre Ablin, Gabriel Peyré, Thomas Moreau
2023 J jnl
Understanding the Regularity of Self-Attention with Optimal Transport.
CoRR
Valérie Castin, Pierre Ablin, Gabriel Peyré
2022 A* conf
A framework for bilevel optimization that enables stochastic and global variance reduction algorithms.
NeurIPS
Mathieu Dagréou, Pierre Ablin, Samuel Vaiter, Thomas Moreau
2022 J jnl
A framework for bilevel optimization that enables stochastic and global variance reduction algorithms.
CoRR
Mathieu Dagréou, Pierre Ablin, Samuel Vaiter, Thomas Moreau
2022 A* conf
Benchopt: Reproducible, efficient and collaborative optimization benchmarks.
NeurIPS
Thomas Moreau, Mathurin Massias, Alexandre Gramfort, Pierre Ablin, Pierre-Antoine Bannier, Benjamin Charlier, Mathieu Dagréou, Tom Dupré la Tour, Ghislain Durif, Cássio F. Dantas, Quentin Klopfenstein, Johan Larsson, En Lai, Tanguy Lefort, Benoît Malézieux, Badr Moufad, Binh T. Nguyen, Alain Rakotomamonjy, Zaccharie Ramzi, Joseph Salmon, Samuel Vaiter
2022 J jnl
Benchopt: Reproducible, efficient and collaborative optimization benchmarks.
CoRR
Thomas Moreau, Mathurin Massias, Alexandre Gramfort, Pierre Ablin, Pierre-Antoine Bannier, Benjamin Charlier, Mathieu Dagréou, Tom Dupré la Tour, Ghislain Durif, Cássio F. Dantas, Quentin Klopfenstein, Johan Larsson, En Lai, Tanguy Lefort, Benoît Malézieux, Badr Moufad, Binh T. Nguyen, Alain Rakotomamonjy, Zaccharie Ramzi, Joseph Salmon, Samuel Vaiter
2022 A* conf
Do Residual Neural Networks discretize Neural Ordinary Differential Equations?
NeurIPS
Michael E. Sander, Pierre Ablin, Gabriel Peyré
2022 J jnl
Do Residual Neural Networks discretize Neural Ordinary Differential Equations?
CoRR
Michael E. Sander, Pierre Ablin, Gabriel Peyré
2022 A conf
Fast and accurate optimization on the orthogonal manifold without retraction.
AISTATS
Pierre Ablin, Gabriel Peyré
2022 A conf
Sinkformers: Transformers with Doubly Stochastic Attention.
AISTATS
Michael E. Sander, Pierre Ablin, Mathieu Blondel, Gabriel Peyré
2021 J jnl
Adaptive Multi-View ICA: Estimation of noise levels for optimal inference.
CoRR
Hugo Richard, Pierre Ablin, Aapo Hyvärinen, Alexandre Gramfort, Bertrand Thirion
2021 J jnl
Fast and accurate optimization on the orthogonal manifold without retraction.
CoRR
Pierre Ablin, Gabriel Peyré
2021 A* conf
Kernel Stein Discrepancy Descent.
ICML
Anna Korba, Pierre-Cyril Aubin-Frankowski, Szymon Majewski, Pierre Ablin
2021 J jnl
Kernel Stein Discrepancy Descent.
CoRR
Anna Korba, Pierre-Cyril Aubin-Frankowski, Szymon Majewski, Pierre Ablin
2021 A* conf
Momentum Residual Neural Networks.
ICML
Michael E. Sander, Pierre Ablin, Mathieu Blondel, Gabriel Peyré
2021 J jnl
Momentum Residual Neural Networks.
CoRR
Michael E. Sander, Pierre Ablin, Mathieu Blondel, Gabriel Peyré
2021 A* conf
Shared Independent Component Analysis for Multi-Subject Neuroimaging.
NeurIPS
Hugo Richard, Pierre Ablin, Bertrand Thirion, Alexandre Gramfort, Aapo Hyvärinen
2021 J jnl
Shared Independent Component Analysis for Multi-Subject Neuroimaging.
CoRR
Hugo Richard, Pierre Ablin, Bertrand Thirion, Alexandre Gramfort, Aapo Hyvärinen
2021 J jnl
Sinkformers: Transformers with Doubly Stochastic Attention.
CoRR
Michael E. Sander, Pierre Ablin, Mathieu Blondel, Gabriel Peyré
2021 J jnl
mvlearn: Multiview Machine Learning in Python.
J. Mach. Learn. Res.
Ronan Perry, Gavin Mischler, Richard Guo, Theo Lee, Alexander Chang, Arman Koul, Cameron Franz, Hugo Richard, Iain Carmichael, Pierre Ablin, Alexandre Gramfort, Joshua T. Vogelstein
2020 J jnl
Deep orthogonal linear networks are shallow.
CoRR
Pierre Ablin
2020 J jnl
Modeling Shared Responses in Neuroimaging Studies through MultiView ICA.
CoRR
Hugo Richard, Luigi Gresele, Aapo Hyvärinen, Bertrand Thirion, Alexandre Gramfort, Pierre Ablin
2020 A* conf
Modeling Shared responses in Neuroimaging Studies through MultiView ICA.
NeurIPS
Hugo Richard, Luigi Gresele, Aapo Hyvärinen, Bertrand Thirion, Alexandre Gramfort, Pierre Ablin
2020 J jnl
Predictive regression modeling with MEG/EEG: from source power to signals and cognitive states.
NeuroImage
David Sabbagh, Pierre Ablin, Gaël Varoquaux, Alexandre Gramfort, Denis A. Engemann
2020 A* conf
Super-efficiency of automatic differentiation for functions defined as a minimum.
ICML
Pierre Ablin, Gabriel Peyré, Thomas Moreau
2020 J jnl
Super-efficiency of automatic differentiation for functions defined as a minimum.
CoRR
Pierre Ablin, Gabriel Peyré, Thomas Moreau
2019 Misc conf
A Quasi-Newton Algorithm on the Orthogonal Manifold for NMF with Transform Learning.
ICASSP
Pierre Ablin, Dylan Fagot, Herwig Wendt, Alexandre Gramfort, Cédric Févotte
2019 B conf
Beyond Pham's algorithm for joint diagonalization.
ESANN
Pierre Ablin, Jean-François Cardoso, Alexandre Gramfort
2019 A* conf
Learning step sizes for unfolded sparse coding.
NeurIPS
Pierre Ablin, Thomas Moreau, Mathurin Massias, Alexandre Gramfort
2019 J jnl
Learning step sizes for unfolded sparse coding.
CoRR
Pierre Ablin, Thomas Moreau, Mathurin Massias, Alexandre Gramfort
2019 A* conf
Manifold-regression to predict from MEG/EEG brain signals without source modeling.
NeurIPS
David Sabbagh, Pierre Ablin, Gaël Varoquaux, Alexandre Gramfort, Denis A. Engemann
2019 J jnl
Manifold-regression to predict from MEG/EEG brain signals without source modeling.
CoRR
David Sabbagh, Pierre Ablin, Gaël Varoquaux, Alexandre Gramfort, Denis A. Engemann
2019 A conf
Stochastic algorithms with descent guarantees for ICA.
AISTATS
Pierre Ablin, Alexandre Gramfort, Jean-François Cardoso, Francis R. Bach
2018 J jnl
A Quasi-Newton algorithm on the orthogonal manifold for NMF with transform learning.
CoRR
Pierre Ablin, Dylan Fagot, Herwig Wendt, Alexandre Gramfort, Cédric Févotte
2018 — conf
Accelerating Likelihood Optimization for ICA on Real Signals.
LVA/ICA
Pierre Ablin, Jean-François Cardoso, Alexandre Gramfort
2018 J jnl
Accelerating likelihood optimization for ICA on real signals.
CoRR
Pierre Ablin, Jean-François Cardoso, Alexandre Gramfort
2018 J jnl
Beyond Pham's algorithm for joint diagonalization.
CoRR
Pierre Ablin, Jean-François Cardoso, Alexandre Gramfort
2018 J jnl
EM algorithms for ICA.
CoRR
Pierre Ablin, Alexandre Gramfort, Jean-François Cardoso, Francis R. Bach
2018 Misc conf
Faster ICA Under Orthogonal Constraint.
ICASSP
Pierre Ablin, Jean-François Cardoso, Alexandre Gramfort
2018 J jnl
Faster Independent Component Analysis by Preconditioning With Hessian Approximations.
IEEE Trans. Signal Process.
Pierre Ablin, Jean-François Cardoso, Alexandre Gramfort
2018 J jnl
Statistical Shape Modeling of the Left Ventricle: Myocardial Infarct Classification Challenge.
IEEE J. Biomed. Health Informatics
Avan Suinesiaputra, Pierre Ablin, Xènia Albà, Martino Alessandrini, Jack Allen, Wenjia Bai, Serkan Çimen, Peter Claes, Brett R. Cowan, Jan D'hooge, Nicolas Duchateau, Jan Ehrhardt, Alejandro F. Frangi, Ali Gooya, Vicente Grau, Karim Lekadir, Allen Lu, Anirban Mukhopadhyay, Ilkay Öksüz, Nripesh Parajuli, Xavier Pennec, Marco Pereañez, Catarina Pinto, Paolo Piras, Marc-Michel Rohé, Daniel Rueckert, Dennis Säring, Maxime Sermesant, Kaleem Siddiqi, Mahdi Tabassian, Luciano Teresi, Sotirios A. Tsaftaris, Matthias Wilms, Alistair A. Young, Xingyu Zhang, Pau Medrano-Gracia
2015 — conf
Detecting Myocardial Infarction Using Medial Surfaces - LV Statistical Modelling Challenge: Myocardial Infarction.
STACOM@MICCAI
Pierre Ablin, Kaleem Siddiqi