Doina Precup

Elaine Lau

Maîtrise recherche - McGill

Jonathan Lebensold

Collaborateur·rice alumni - McGill

Sitao Luan

Doctorat - McGill

Ray Luo

Doctorat - McGill

Superviseur⋅e principal⋅e :

G McCracken

Doctorat - McGill

Jaume Minano Masip

Doctorat - McGill

Shahrad Mohammadzadeh

Maîtrise recherche - McGill

Superviseur⋅e principal⋅e :

Gabriela Moisescu-Pareja

Maîtrise recherche - McGill

Girdhar Neil Girdhar

Collaborateur·rice de recherche - McGill

Padideh Nouri

Maîtrise recherche - UdeM

Charles Onu

Doctorat - McGill

Doctorat - McGill

Co-superviseur⋅e :

Nate Rahn

Doctorat - McGill

Superviseur⋅e principal⋅e :

Marc Gendron-Bellemare

Janarthanan Rajendran

Collaborateur·rice alumni - UdeM

Superviseur⋅e principal⋅e :

Sarath Chandar

Sahand Rezaei-Shoshtari

Doctorat - McGill

Co-superviseur⋅e :

Doctorat - McGill

Co-superviseur⋅e :

Blake Richards

samiemandana@gmail.com

Doctorat - McGill

Maîtrise recherche - McGill

Nishanth Anand Vemgal

Doctorat - McGill

Co-superviseur⋅e :

Doctorat - McGill

Doctorat - McGill

Co-superviseur⋅e :

Samira Ebrahimi Kahou

Stagiaire de recherche - McGill

Keyu Wang

Stagiaire de recherche - McGill

Skipper : combiner l’abstraction spatiale et temporelle afin d’améliorer la généralisation

Steve Wen

Baccalauréat - McGill

Shuyuan Zhang

Doctorat - McGill

Harry Zhao

Doctorat - McGill

Co-superviseur⋅e :

Billets de blogue

Generic thumbnail for Mila Blog articles.

22 février 2024

par

Mingde Harry Zhao

Safa Alver

Harm van Seijen

Romain Laroche

Doina Precup

Yoshua Bengio

Lire l'article

Publications

Training Language Models to Self-Correct via Reinforcement Learning

Aviral Kumar

Vincent Zhuang

Rishabh Agarwal

Yi Su

John D Co-Reyes

Avi Singh

Kate Baumli

Shariq N Iqbal

Colton Bishop

Rebecca Roelofs

Lei M Zhang

Kay McKinney

Disha Shrivastava

Cosmin Paduraru

George Tucker

Feryal Behbahani

Aleksandra Faust

Self-correction is a highly desirable capability of large language models (LLMs), yet it has consistently been found to be largely ineffecti… (voir plus)ve in modern LLMs. Existing approaches for training self-correction either require multiple models or rely on a more capable model or other forms of supervision. To this end, we develop a multi-turn online reinforcement learning (RL) approach, SCoRe, that significantly improves an LLM's self-correction ability using entirely self-generated data. To build SCoRe, we first show that variants of supervised fine-tuning (SFT) on offline model-generated correction traces are insufficient for instilling self-correction behavior. In particular, we observe that training via SFT either suffers from a distribution mismatch between the training data and the model's own responses or implicitly prefers only a certain mode of correction behavior that is often not effective at test time. SCoRe addresses these challenges by training under the model's own distribution of self-generated correction traces and using appropriate regularization to steer the learning process into learning a self-correction strategy that is effective at test time as opposed to simply fitting high-reward responses for a given prompt. This regularization prescribes running a first phase of RL on a base model to generate a policy initialization that is less susceptible to collapse and then using a reward bonus to amplify self-correction during training. When applied to Gemini 1.0 Pro and 1.5 Flash models, we find that SCoRe achieves state-of-the-art self-correction performance, improving the base models' self-correction by 15.6% and 9.1% respectively on the MATH and HumanEval benchmarks.

2024-09-19

ArXiv (prépublication)

An Attentive Approach for Building Partial Reasoning Agents from Pixels

Safa Alver

We study the problem of building reasoning agents that are able to generalize in an effective manner. Towards this goal, we propose an end-t… (voir plus)o-end approach for building model-based reinforcement learning agents that dynamically focus their reasoning to the relevant aspects of the environment: after automatically identifying the distinct aspects of the environment, these agents dynamically filter out the relevant ones and then pass them to their simulator to perform partial reasoning. Unlike existing approaches, our approach works with pixel-based inputs and it allows for interpreting the focal points of the agent. Our quantitative analyses show that the proposed approach allows for effective generalization in high-dimensional domains with raw observational inputs. We also perform ablation analyses to validate of design choices. Finally, we demonstrate through qualitative analyses that our approach actually allows for building agents that focus their reasoning on the relevant aspects of the environment.

2024-09-17

TMLR (accepté)

An Attentive Approach for Building Partial Reasoning Agents from Pixels

Safa Alver

We study the problem of building reasoning agents that are able to generalize in an effective manner. Towards this goal, we propose an end-t… (voir plus)o-end approach for building model-based reinforcement learning agents that dynamically focus their reasoning to the relevant aspects of the environment: after automatically identifying the distinct aspects of the environment, these agents dynamically filter out the relevant ones and then pass them to their simulator to perform partial reasoning. Unlike existing approaches, our approach works with pixel-based inputs and it allows for interpreting the focal points of the agent. Our quantitative analyses show that the proposed approach allows for effective generalization in high-dimensional domains with raw observational inputs. We also perform ablation analyses to validate our design choices. Finally, we demonstrate through qualitative analyses that our approach actually allows for building agents that focus their reasoning on the relevant aspects of the environment.

2024-09-17

TMLR (accepté)

A Look at Value-Based Decision-Time vs. Background Planning Methods Across Different Settings

Safa Alver

In model-based reinforcement learning (RL), an agent can leverage a learned model to improve its way of behaving in different ways. Two of t… (voir plus)he prevalent ways to do this are through decision-time and background planning methods. In this study, we are interested in understanding how the value-based versions of these two planning methods will compare against each other across different settings. Towards this goal, we first consider the simplest instantiations of value-based decision-time and background planning methods and provide theoretical results on which one will perform better in the regular RL and transfer learning settings. Then, we consider the modern instantiations of them and provide hypotheses on which one will perform better in the same settings. Finally, we perform illustrative experiments to validate these theoretical results and hypotheses. Overall, our findings suggest that even though value-based versions of the two planning methods perform on par in their simplest instantiations, the modern instantiations of value-based decision-time planning methods can perform on par or better than the modern instantiations of value-based background planning methods in both the regular RL and transfer learning settings.

2024-08-01

EWRL/2024/Workshop (accepté)

The Heterophilic Graph Learning Handbook: Benchmarks, Models, Theoretical Analysis, Applications and Challenges

Sitao Luan

Chenqing Hua

Qincheng Lu

Liheng Ma

Lirong Wu

Xinyu Wang

Minkai Xu

Xiao-Wen Chang

Rex Ying

Stan Z. Li

Jian Tang

Guy Wolf

Stefanie Jegelka

Homophily principle, \ie{} nodes with the same labels or similar attributes are more likely to be connected, has been commonly believed to b… (voir plus)e the main reason for the superiority of Graph Neural Networks (GNNs) over traditional Neural Networks (NNs) on graph-structured data, especially on node-level tasks. However, recent work has identified a non-trivial set of datasets where GNN's performance compared to the NN's is not satisfactory. Heterophily, i.e. low homophily, has been considered the main cause of this empirical observation. People have begun to revisit and re-evaluate most existing graph models, including graph transformer and its variants, in the heterophily scenario across various kinds of graphs, e.g. heterogeneous graphs, temporal graphs and hypergraphs. Moreover, numerous graph-related applications are found to be closely related to the heterophily problem. In the past few years, considerable effort has been devoted to studying and addressing the heterophily issue. In this survey, we provide a comprehensive review of the latest progress on heterophilic graph learning, including an extensive summary of benchmark datasets and evaluation of homophily metrics on synthetic graphs, meticulous classification of the most updated supervised and unsupervised learning methods, thorough digestion of the theoretical analysis on homophily/heterophily, and broad exploration of the heterophily-related applications. Notably, through detailed experiments, we are the first to categorize benchmark heterophilic datasets into three sub-categories: malignant, benign and ambiguous heterophily. Malignant and ambiguous datasets are identified as the real challenging datasets to test the effectiveness of new models on the heterophily challenge. Finally, we propose several challenges and future directions for heterophilic graph representation learning.

2024-07-12

ArXiv (prépublication)

Functional Acceleration for Policy Mirror Descent

Veronica Chelu

2024-06-19

ICML.cc/2024/Workshop/ARLET (poster)

Functional Acceleration for Policy Mirror Descent

Veronica Chelu

We apply functional acceleration to the Policy Mirror Descent (PMD) general family of algorithms, which cover a wide range of novel and fund… (voir plus)amental methods in Reinforcement Learning (RL). Leveraging duality, we propose a momentum-based PMD update. By taking the functional route, our approach is independent of the policy parametrization and applicable to large-scale optimization, covering previous applications of momentum at the level of policy parameters as a special case. We theoretically analyze several properties of this approach and complement with a numerical ablation study, which serves to illustrate the policy optimization dynamics on the value polytope, relative to different algorithmic design choices in this space. We further characterize numerically several features of the problem setting relevant for functional acceleration, and lastly, we investigate the impact of approximation on their learning mechanics.

2024-06-19

ICML.cc/2024/Workshop/ARLET (poster)

QGFN: Controllable Greediness with Action Values

Elaine Lau

Stephen Zhewen Lu

Ling Pan

Emmanuel Bengio

Generative Flow Networks (GFlowNets; GFNs) are a family of reward/energy-based generative methods for combinatorial objects, capable of gene… (voir plus)rating diverse and high-utility samples. However, biasing GFNs towards producing high-utility samples is non-trivial. In this work, we leverage connections between GFNs and reinforcement learning (RL) and propose to combine the GFN policy with an action-value estimate,

2024-06-17

ICML.cc/2024/Workshop/SPIGM (poster)

Recurrent Policies Are Not Enough for Continual Reinforcement Learning

Nathan Samuel de Lara

Veronica Chelu

Continual Reinforcement Learning (CRL) aims to develop algorithms that adapt to non-stationary sequences of tasks. A promising recent approa… (voir plus)ch utilizes Recurrent Neural Networks (RNNs) to learn contextual Markov Decision Process (MDP) embeddings. This enables a reinforcement learning (RL) agent to discern the optimality of actions across diverse tasks. In this study, we examine two critical failure modes in the learning of these contextual MDP embeddings. Specifically, we find that RNNs are prone to catastrophic forgetting, manifesting in two distinct ways: (i) embedding collapse---where agents initially learn a contextual task structure that later collapses to a single task, and (ii) embedding drift---where learning embeddings for new MDPs interferes with embeddings the RNN outputs for previous MDPs in the sequence, leading to suboptimal performance of downstream policy networks conditioned on stale embeddings. We explore the effects of various objective functions and network architectures concerning these failure modes, revealing that one of these modes consistently emerges across different setups.

2024-06-07

rl-conference.cc/RLC/2024/Workshop/ICBINB (poster)

On the Limits of Multi-modal Meta-Learning with Auxiliary Task Modulation Using Conditional Batch Normalization

Jordi Armengol-Estap'e

Vincent Michalski

Ramnath Kumar

Pierre-Luc St-Charles

Samira Ebrahimi Kahou

Few-shot learning aims to learn representations that can tackle novel tasks given a small number of examples. Recent studies show that cross… (voir plus)-modal learning can improve representations for few-shot classification. More specifically, language is a rich modality that can be used to guide visual learning. In this work, we experiment with a multi-modal architecture for few-shot learning that consists of three components: a classifier, an auxiliary network, and a bridge network. While the classifier performs the main classification task, the auxiliary network learns to predict language representations from the same input, and the bridge network transforms high-level features of the auxiliary network into modulation parameters for layers of the few-shot classifier using conditional batch normalization. The bridge should encourage a form of lightweight semantic alignment between language and vision which could be useful for the classifier. However, after evaluating the proposed approach on two popular few-shot classification benchmarks we find that a) the improvements do not reproduce across benchmarks, and b) when they do, the improvements are due to the additional compute and parameters introduced by the bridge network. We contribute insights and recommendations for future work in multi-modal meta-learning, especially when using language representations.

2024-05-29

ArXiv (prépublication)

Partial Models for Building Adaptive Model-Based Reinforcement Learning Agents

Safa Alver

Ali Rahimi-Kalahroudi

In neuroscience, one of the key behavioral tests for determining whether a subject of study exhibits model-based behavior is to study its ad… (voir plus)aptiveness to local changes in the environment. In reinforcement learning, however, recent studies have shown that modern model-based agents display poor adaptivity to such changes. The main reason for this is that modern agents are typically designed to improve sample efficiency in single task settings and thus do not take into account the challenges that can arise in other settings. In local adaptation settings, one particularly important challenge is in quickly building and maintaining a sufficiently accurate model after a local change. This is challenging for deep model-based agents as their models and replay buffers are monolithic structures lacking distribution shift handling capabilities. In this study, we show that the conceptually simple idea of partial models can allow deep model-based agents to overcome this challenge and thus allow for building locally adaptive model-based agents. By modeling the different parts of the state space through different models, the agent can not only maintain a model that is accurate across the state space, but it can also quickly adapt it in the presence of a local change in the environment. We demonstrate this by showing that the use of partial models in agents such as deep Dyna-Q, PlaNet and Dreamer can allow for them to effectively adapt to the local changes in their environments.

2024-05-27

ArXiv (prépublication)

More Efficient Randomized Exploration for Reinforcement Learning via Approximate Sampling

Haque Ishfaq

Yixin Tan

Yu Yang

Qingfeng Lan

Jianfeng Lu

A. Rupam Mahmood

Pan Xu

2024-05-14

rl-conference.cc/RLC/2024/Conference (publié)