Aaron Courville

Anirudh Buvanesh

Doctorat - UdeM

Superviseur⋅e principal⋅e :

Laurent Charlin

anirudb1102@gmail.com

Razvan Ciuca

Maîtrise recherche - Université de Montréal

Juan Duque

Doctorat - UdeM

Doctorat - UdeM

Doctorat - UdeM

Amr Khalifa

Doctorat - UdeM

Samuel Lavoie

Doctorat - UdeM

Zhixuan Lin

Doctorat - UdeM

Ahmed Masry

Collaborateur·rice de recherche - N/A

Doctorat - UdeM

Superviseur⋅e principal⋅e :

Doctorat - UdeM

Superviseur⋅e principal⋅e :

Doctorat - UdeM

Doctorat - UdeM

Co-superviseur⋅e :

andrei.nicolicioiu@gmail.com

Andrei Nicolicioiu

Doctorat - UdeM

Michell Mercedes Payano Perez

Google Scholar

Evgenii Nikishin

Collaborateur·rice alumni - UdeM

Superviseur⋅e principal⋅e :

Doctorat - UdeM

Johan Samir Obando Ceron

Doctorat - UdeM

Co-superviseur⋅e :

Stagiaire de recherche - UdeM

Dereck Piché

Maîtrise recherche - UdeM

pichedereck@gmail.com

Esra'a Saleh

Doctorat - UdeM

Superviseur⋅e principal⋅e :

Glen Berseth

Google Scholar

Vedant Shah

Doctorat - UdeM

Maîtrise recherche - UdeM

Superviseur⋅e principal⋅e :

Anna (Cheng-Zhi) Huang

Doctorat - UdeM

Yusong Wu

Doctorat - UdeM

Superviseur⋅e principal⋅e :

Anna (Cheng-Zhi) Huang

sujin yun

Doctorat - UdeM

Doctorat - UdeM

Doctorat - UdeM

Co-superviseur⋅e :

Yoshua Bengio

Hattie Zhou

Doctorat - UdeM

Superviseur⋅e principal⋅e :

Hugo Larochelle

Publications

Improving Compositional Generalization using Iterated Learning and Simplicial Embeddings

Yi Ren

Samuel Lavoie

Mikhail Galkin

Danica J. Sutherland

Language Model Alignment with Elastic Reset

Michael Noukhovitch

Samuel Lavoie

Florian Strub

Finetuning language models with reinforcement learning (RL), e.g. from human feedback (HF), is a prominent method for alignment. But optimiz… (voir plus)ing against a reward model can improve on reward while degrading performance in other areas, a phenomenon known as reward hacking, alignment tax, or language drift. First, we argue that commonly-used test metrics are insufficient and instead measure how different algorithms tradeoff between reward and drift. The standard method modified the reward with a Kullback-Lieber (KL) penalty between the online and initial model. We propose Elastic Reset, a new algorithm that achieves higher reward with less drift without explicitly modifying the training objective. We periodically reset the online model to an exponentially moving average (EMA) of itself, then reset the EMA model to the initial model. Through the use of an EMA, our model recovers quickly after resets and achieves higher reward with less drift in the same number of steps. We demonstrate that fine-tuning language models with Elastic Reset leads to state-of-the-art performance on a small scale pivot-translation benchmark, outperforms all baselines in a medium-scale RLHF-like IMDB mock sentiment task and leads to a more performant and more aligned technical QA chatbot with LLaMA-7B. Code available at github.com/mnoukhov/elastic-reset.

Let the Flows Tell: Solving Graph Combinatorial Problems with GFlowNets

Hanjun Dai

Ling Pan

Versatile Energy-Based Probabilistic Models for High Energy Physics

Taoli Cheng

Discovering the Electron Beam Induced Transition Rates for Silicon Dopants in Graphene with Deep Neural Networks in the STEM

Kevin M Roccapriore

Max Schwarzer

Joshua Greaves

Jesse Farebrother

Colton Bishop

Maxim Ziatdinov

Igor Mordatch

Ekin Dogus Cubuk

Pablo Samuel Castro

Marc Gendron-Bellemare

Sergei V Kalinin

2023-07-22

Microscopy and Microanalysis (publié)

Meta-Value Learning: a General Framework for Learning with Learning Awareness

Tim Cooijmans

Milad Aghajohari

2023-07-17

ArXiv (preprint)

Bigger, Better, Faster: Human-level Atari with human-level efficiency

Max Schwarzer

Johan Samir Obando Ceron

Marc Gendron-Bellemare

Pablo Samuel Castro

We introduce a value-based RL agent, which we call BBF, that achieves super-human performance in the Atari 100K benchmark. BBF relies on sca… (voir plus)ling the neural networks used for value estimation, as well as a number of other design choices that enable this scaling in a sample-efficient manner. We conduct extensive analyses of these design choices and provide insights for future work. We end with a discussion about updating the goalposts for sample-efficient RL research on the ALE. We make our code and data publicly available at https://github.com/google-research/google-research/tree/master/bigger_better_faster.

2023-07-03

Proceedings of the 40th International Conference on Machine Learning (publié)

Learning with Learning Awareness using Meta-Values

Tim Cooijmans

Milad Aghajohari

2023-06-19

ICML.cc/2023/Workshop/Frontiers4LCD (publié)

Bigger, Better, Faster: Human-level Atari with human-level efficiency

Max Schwarzer

Johan Samir Obando Ceron

Marc Gendron-Bellemare

Pablo Samuel Castro

2023-05-30

ArXiv (prépublication)

arxiv.org

Let the Flows Tell: Solving Graph Combinatorial Optimization Problems with GFlowNets

Hanjun Dai

Ling Pan

Combinatorial optimization (CO) problems are often NP-hard and thus out of reach for exact algorithms, making them a tempting domain to appl… (voir plus)y machine learning methods. The highly structured constraints in these problems can hinder either optimization or sampling directly in the solution space. On the other hand, GFlowNets have recently emerged as a powerful machinery to efficiently sample from composite unnormalized densities sequentially and have the potential to amortize such solution-searching processes in CO, as well as generate diverse solution candidates. In this paper, we design Markov decision processes (MDPs) for different combinatorial problems and propose to train conditional GFlowNets to sample from the solution space. Efficient training techniques are also developed to benefit long-range credit assignment. Through extensive experiments on a variety of different CO tasks with synthetic and realistic data, we demonstrate that GFlowNet policies can efficiently find high-quality solutions. Our implementation is open-sourced at https://github.com/zdhNarsil/GFlowNet-CombOpt.

2023-05-26

ArXiv (prépublication)

arxiv.org

Mastering the Unsupervised Reinforcement Learning Benchmark from Pixels

Sai Rajeswar

Pietro Mazzaglia

Tim Verbelen

Alexandre Piché

Bart Dhoedt

Alexandre Lacoste

Controlling artificial agents from visual sensory data is an arduous task. Reinforcement learning (RL) algorithms can succeed but require la… (voir plus)rge amounts of interactions between the agent and the environment. To alleviate the issue, unsupervised RL proposes to employ self-supervised interaction and learning, for adapting faster to future tasks. Yet, as shown in the Unsupervised RL Benchmark (URLB; Laskin et al. 2021), whether current unsupervised strategies can improve generalization capabilities is still unclear, especially in visual control settings. In this work, we study the URLB and propose a new method to solve it, using unsupervised model-based RL, for pre-training the agent, and a task-aware fine-tuning strategy combined with a new proposed hybrid planner, Dyna-MPC, to adapt the agent for downstream tasks. On URLB, our method obtains 93.59% overall normalized performance, surpassing previous baselines by a staggering margin. The approach is empirically evaluated through a large-scale empirical study, which we use to validate our design choices and analyze our models. We also show robust performance on the Real-Word RL benchmark, hinting at resiliency to environment perturbations during adaptation. Project website: https://masteringurlb.github.io/

2023-04-24

ICML.cc/2023/Conference (poster)

proceedings.mlr.press

Mastering the Unsupervised Reinforcement Learning Benchmark from Pixels

Sai Rajeswar

Pietro Mazzaglia

Tim Verbelen

Alexandre Piché

Bart Dhoedt

Alexandre Lacoste

2023-04-24

ICML.cc/2023/Conference (publié)