Source Themes

CrossSplit: Mitigating Label Noise Memorization through Data Splitting
ICML, 2023
J. Kim, A. Baratin, Y. Zhang, S. Lacoste-Julien
arXiv
Promoting Exploration in Memory-Augmented Adam using Critical Momenta
arXiv, 2023
P. Malviya, G. Mordido, A. Baratin, R. Babanezhad Harikandeh, J. Huang, S. Lacoste-Julien, R. Pascanu, S. Chandar
arXiv
Lazy vs hasty: linearization in deep networks impacts learning schedule based on example difficulty
TMLR, 2022
T. George, G. Lajoie, A. Baratin
arXiv
Expressiveness and Learnability: A Unifying View for Evaluating Self-Supervised Learning
arxiv, 2022
Y. Lu, Z. Liu, A. Baratin, R. Laroche, A. Courville, A. Sordoni
PDF
Implicit Regularization via Neural Feature Alignment
AISTATS 2021
Aristide Baratin, Thomas George, Cesar Laurent, R Devon Hjelm, Guillaume Lajoie, Pascal Vincent, Simon Lacoste-Julien
Code Video arXiv
On the Regularity of Attention
Note: This is a conference version of “A Mathematical Theory of Attention” arXiv:2007.02876.
James Vuckovic, Aristide Baratin, Remi Tachet des Combes
arXiv press
A Mathematical Theory of Attention
arxiv, 2021
James Vuckovic, Aristide Baratin, Remi Tachet des Combes
arXiv
On the Spectral Bias of Neural Networks
ICML 2019
Nasim Rahaman, Aristide Baratin, Devansh Arpit, Felix Draxler, Min Lin, Fred A. Hamprecht, Yoshua Bengio, Aaron Courville
Code Slides arXiv
A Modern Take on the Bias-Variance Tradeoff in Neural Networks
ICML 2019, Workshop on Identifying and Understanding Deep Learning Phenomena
Brady Neal, Sarthak Mittal, Aristide Baratin, Vinayak Tantia, Matthew Scicluna, Simon Lacoste-Julien, Ioannis Mitliagkas
arXiv
MINE: Mutual Information Neural Estimation
ICML 2018
Mohamed Ishmael Belghazi, Aristide Baratin, Sai Rajeswar, Sherjil Ozair, Yoshua Bengio, Aaron Courville, R Devon Hjelm
Video arXiv