About
Hi, I'm Gaspard Lambrechts, a PhD student doing research in reinforcement learning under the supervision of Damien Ernst at the University of Liège. My research focuses on reinforcement learning in partially observable environments.
Publications
-
Behind the Myth of Exploration in Policy Gradients.
Adrien Bolland, Gaspard Lambrechts, Damien Ernst.
Preprint, January 2024. Article. -
Warming Up Recurrent Neural Networks to Maximize Reachable Multistability Greatly Improves Learning.
Gaspard Lambrechts, Florent De Geeter, Nicolas Vecoven, Damien Ernst, Guillaume Drion.
In Neural Networks, Augustus 2023. Article. Blog Post. -
Informed POMDP: Leveraging Additional Information in Model-Based RL.
Gaspard Lambrechts, Adrien Bolland, Damien Ernst.
In ICML Workshop on New Frontiers in Learning, Control, and Dynamical Systems, July 2023. Article. Poster. -
Belief States of POMDPs and Internal States of Recurrent RL Agents: an Empirical Analysis of their Mutual Information.
Gaspard Lambrechts, Adrien Bolland, Damien Ernst.
In European Workshop on Reinforcement Learning, September 2022. Article. Poster. -
Recurrent Networks, Hidden States and Beliefs in Partially Observable Environments.
Gaspard Lambrechts, Adrien Bolland, Damien Ernst.
In Transaction on Machine Learning Research, Augustus 2022. Article. Blog Post.
Talks
-
Learning to Memorize the Past by Learning to Predict the Future
VUB Reinforcement Learning Talks, November 17th, 2023. Slides.
Teaching
Code
Contact
Gaspard Lambrechts, I.106Montefiore Institute, B28
Allée de la découverte, 10
4000 Liège
Belgium