Biography
Pierre Wolinski is Associate Professor of Computer Science at Paris Dauphine University - PSL since September 2024. His research is in the theory of neural networks, including their optimization and regularization. He teaches in several areas of artificial intelligence, such as data analysis, machine learning, and deep learning. Previously, he did his Ph.D. on neural network pruning and hyperparameter removal, and several post-docs focused on neural network initialization and second-order optimization.
Career:
- 2016-2020: Ph.D. in Computer Science (Paris-Saclay University, Inria Saclay), supervised by Guillaume Charpiat and Yann Ollivier.
- 2020-2021: post-doc (University of Oxford), supervised by Judith Rousseau.
- 2021-2023: post-doc (Inria Grenoble-Alpes, UGA), supervised by Julyan Arbel.
- 2023-2024: post-doc (Institut de Mathématiques d'Orsay, Paris-Saclay Univ.), supervised by Gilles Blanchard and Christophe Giraud.
Publications
Communications avec actes
Arbel M., Menegaux R., WOLINSKI P. (2023), Rethinking Gauss-Newton for learning over-parameterized models, in A. Oh ; T. Naumann ; A. Globerson ; K. Saenko ; M. Hardt ; S. Levine, Neural Information Processing Systems Foundation, Inc., 33379-33402 p. 
Blier L., WOLINSKI P., Ollivier Y. (2019), Learning with Random Learning Rates, in Ulf Brefeld ; Elisa Fromont ; Andreas Hotho ; Arno Knobbe ; Marloes Maathuis ; Céline Robardet, Berlin Heidelberg, Springer International Publishing, 449-464 p.
Communications sans actes
Wolinski P., Arbel J. (2025), How Can the Pre-Activations Remain Gaussian as They Propagate through a Neural Network?, European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, Nectar track, Porto, Portugal
WOLINSKI P., Arbel J. (2022), Imposing Gaussian Pre-Activations in a Neural Network, 53e journées des statistiques, Lyon, France
WOLINSKI P., Charpiat G., Ollivier Y. (2022), An Equivalence between Bayesian Priors and Penalties in Variational Inference, International Society for Bayesian Analysis, Montréal, Canada