Yann Ollivier

Research Scientist

After a PhD in probability and group theory, I joined the CNRS, the French national research institute. Initially I worked on unveiling connections between probability, Markov chains, differential geometry and discrete geometry. For this work I was awarded the bronze medal of the CNRS in 2011.

I’ve had a lifelong interest in artificial intelligence, and around 2010 I decided to focus my research on machine learning. At that point I joined the computer science department at Paris-Sud university. Following the industrial development of deep learning, in 2017 I joined Facebook AI Research.
In the long term, I am interested in building general artificial intelligence systems. In the shorter term I work on understanding and improving the learning algorithms for neural networks. More specific fields of interest include the geometry of gradient descent algorithms, the dynamics of recurrent networks and online learning, better algorithms for reinforcement learning, and what “learning” means in terms of information theory.

Related Links

Personal Website

Latest Publications

ICML - June 10, 2019

Separating value functions across time-scales

Joshua Romoff, Peter Henderson, Ahmed Touati, Emma Brunskill, Joelle Pineau, Yann Ollivier

ICML - June 10, 2019

Making Deep Q-learning Methods Robust to Time Discretization

Corentin Tallec, Léonard Blier, Yann Ollivier

ICML - June 9, 2019

First-order Adversarial Vulnerability of Neural Networks and Input Dimension

Carl-Johann Simon-Gabriel, Yann Ollivier, Bernhard Scholkopf, Léon Bottou, David Lopez-Paz

NeurIPS 2018 - December 2, 2018

The Description Length of Deep Learning Models

Léonard Blier, Yann Ollivier

ICML 2018 - July 9, 2018

Mixed batches and symmetric discriminators for GAN training

Thomas Lucas, Corentin Tallec, Jakob Verbeek, Yann Ollivier

ICLR 2018 - April 30, 2018

Can Recurrent Neural Networks Wrap Time?

Corentin Tallec, Yann Ollivier