Aristide Baratin

About Aristide Baratin

Aristide Baratin, With an exceptional h-index of 17 and a recent h-index of 13 (since 2020), a distinguished researcher at Université de Montréal, specializes in the field of machine learning, deep learning, theoretical physics.

His recent articles reflect a diverse array of research interests and contributions to the field:

Maxwell's Demon at Work: Efficient Pruning by Leveraging Saturation of Neurons

Unsupervised Concept Discovery Mitigates Spurious Correlations

How connectivity structure shapes rich and lazy learning in neural circuits

Lookbehind optimizer: k steps back, 1 step forward

Promoting Exploration in Memory-Augmented Adam using Critical Momenta

CrossSplit: Mitigating Label Noise Memorization through Data Splitting

Using Representation Expressiveness and Learnability to Evaluate Self-Supervised Learning Methods

Some phenomenological investigations in deep learning

Aristide Baratin Information

University

Position

Mila

Citations(all)

3934

Citations(since 2020)

3174

Cited By

1604

hIndex(all)

17

hIndex(since 2020)

13

i10Index(all)

19

i10Index(since 2020)

15

Email

University Profile Page

Google Scholar

Aristide Baratin Skills & Research Interests

machine learning

deep learning

theoretical physics

Top articles of Aristide Baratin

Maxwell's Demon at Work: Efficient Pruning by Leveraging Saturation of Neurons

arXiv preprint arXiv:2403.07688

2024/3/12

Evgenii Nikishin
Evgenii Nikishin

H-Index: 1

Aristide Baratin
Aristide Baratin

H-Index: 12

Unsupervised Concept Discovery Mitigates Spurious Correlations

arXiv preprint arXiv:2402.13368

2024/2/20

How connectivity structure shapes rich and lazy learning in neural circuits

ArXiv

2023/10/12

Lookbehind optimizer: k steps back, 1 step forward

arXiv preprint arXiv:2307.16704

2023/7/31

Pranshu Malviya
Pranshu Malviya

H-Index: 0

Aristide Baratin
Aristide Baratin

H-Index: 12

Promoting Exploration in Memory-Augmented Adam using Critical Momenta

arXiv preprint arXiv:2307.09638

2023/7/18

CrossSplit: Mitigating Label Noise Memorization through Data Splitting

2023/7

Using Representation Expressiveness and Learnability to Evaluate Self-Supervised Learning Methods

Transactions on Machine Learning Research

2023/6/28

Some phenomenological investigations in deep learning

2022/10/26

Aristide Baratin
Aristide Baratin

H-Index: 12

Lazy vs hasty: linearization in deep networks impacts learning schedule based on example difficulty

arXiv preprint arXiv:2209.09658

2022/9/19

Implicit Regularization via Neural Feature Alignment

2021/3/18

On the regularity of attention

arXiv preprint arXiv:2102.05628

2021/2/10

Aristide Baratin
Aristide Baratin

H-Index: 12

A mathematical theory of attention

arXiv preprint arXiv:2007.02876

2020/7/6

Aristide Baratin
Aristide Baratin

H-Index: 12

See List of Professors in Aristide Baratin University(Université de Montréal)

Co-Authors

academic-engine