Portrait de Marc Gendron-Bellemare n'est pas disponible

Marc Gendron-Bellemare

Membre industriel principal
Chaire en IA Canada-CIFAR
Professeur adjoint, McGill University, École d'informatique
Professeur asssocié, Université de Montréal, Département d'informatique et de recherche opérationnelle
Directeur scientifique, Reliant AI
Sujets de recherche
Apprentissage de représentations
Apprentissage par renforcement
Grands modèles de langage (LLM)

Biographie

J'occupe actuellement le poste de directeur scientifique à Reliant AI. Je suis également professeur adjoint à l'École d'informatique de l'Université McGill et professeur adjoint au Département d'informatique et de recherche opérationnelle (DIRO) de l'Université de Montréal.

Précédemment, j'ai travaillé à Google Brain à Montréal, où je me concentrais sur l'apprentissage par renforcement. De 2013 à 2017, j'ai travaillé chez DeepMind au Royaume-Uni. J'ai obtenu un doctorat de l'Université de l'Alberta en travaillant avec Michael Bowling et Joel Veness.

Ma recherche se situe au carrefour de l'apprentissage par renforcement et de la prédiction probabiliste. Je m'intéresse aussi à l'apprentissage profond, à la modélisation générative, à l'apprentissage en ligne et à la théorie de l'information.

Étudiants actuels

Doctorat - McGill
Co-superviseur⋅e :
Doctorat - McGill
Co-superviseur⋅e :
Doctorat - UdeM
Superviseur⋅e principal⋅e :
Doctorat - McGill
Superviseur⋅e principal⋅e :

Publications

A Geometric Perspective on Optimal Representations for Reinforcement Learning
Will Dabney
Robert Dadashi
Dale Schuurmans
Tor Lattimore
Clare Lyle
Dopamine: A Research Framework for Deep Reinforcement Learning
Subhodeep Moitra
Carles Gelada
Saurabh Kumar
Deep reinforcement learning (deep RL) research has grown significantly in recent years. A number of software offerings now exist that provid… (voir plus)e stable, comprehensive implementations for benchmarking. At the same time, recent deep RL research has become more diverse in its goals. In this paper we introduce Dopamine, a new research framework for deep RL that aims to support some of that diversity. Dopamine is open-source, TensorFlow-based, and provides compact and reliable implementations of some state-of-the-art deep RL agents. We complement this offering with a taxonomy of the different research objectives in deep RL research. While by no means exhaustive, our analysis highlights the heterogeneity of research in the field, and the value of frameworks such as ours.
Approximate Exploration through State Abstraction
Although exploration in reinforcement learning is well understood from a theoretical point of view, provably correct methods remain impracti… (voir plus)cal. In this paper we study the interplay between exploration and approximation, what we call approximate exploration. Our main goal is to further our theoretical understanding of pseudo-count based exploration bonuses (Bellemare et al., 2016), a practical exploration scheme based on density modelling. As a warm-up, we quantify the performance of an exploration algorithm, MBIE-EB (Strehl and Littman, 2008), when explicitly combined with state aggregation. This allows us to confirm that, as might be expected, approximation allows the agent to trade off between learning speed and quality of the learned policy. Next, we show how a given density model can be related to an abstraction and that the corresponding pseudo-count bonus can act as a substitute in MBIE-EB combined with this abstraction, but may lead to either under- or over-exploration. Then, we show that a given density model also defines an implicit abstraction, and find a surprising mismatch between pseudo-counts derived either implicitly or explicitly. Finally we derive a new pseudo-count bonus alleviating this issue.