Vivien Cabannes
POSTDOCTORAL RESEARCHER | NEW YORK CITY, UNITED STATES
Vivien Cabannes is a researcher in core machine learning. His interests include learning with few human annotations (weakly supervised learning, representation learning, active learning), as well as understanding the inner workings of deployed machine learning methods.
Research Areas
Vivien's Publications
May 27, 2026
OPEN SOURCE
AutoformBot: Formalizing Mathematics at Scale
Amaury Hayat, Ahmad Rammal, Charles Arnal, Julia Kempe, Niket Patel, Remi Munos, Vivien Cabannes
May 27, 2026
July 24, 2024
COMPUTER VISION
X-Sample Contrastive Loss: Improving Contrastive Learning with Sample Similarity Graphs
Randall Balestriero, Diane Bouchacourt, Kyunghyun Cho, Mark Ibrahim, Pietro Astolfi, Vivien Cabannes, Vlad Sobal, Yann LeCun
July 24, 2024
October 05, 2023
CORE MACHINE LEARNING
Active Self-Supervised Learning: A Few Low-Cost Relationships Are All You Need
Vivien Cabannes, Leon Bottou, Yann LeCun, Randall Balestriero
October 05, 2023
June 26, 2023
CORE MACHINE LEARNING
The SSL Interplay: Augmentations, Inductive Bias, and Generalization
Bobak Kiani, Vivien Cabannes, Alberto Bietti, Randall Balestriero, Yann LeCun
June 26, 2023
Our approach
Latest news
Foundational models
Meta © 2026