Irina Rish

Ivan Anokhin

Doctorat - UdeM

Co-superviseur⋅e :

Samira Ebrahimi Kahou

ivan.anokhin@mila.quebec

Doctorat - UdeM

Arjun Ashok

Doctorat

Co-superviseur⋅e :

Alexandre Drouin

arjun.ashok@mila.quebec

Reza Bayat

Maîtrise recherche - UdeM

Co-superviseur⋅e :

Pouya Bashivan

reza.bayat@mila.quebec

rishika.bhagwatkar@mila.quebec

Rishika Bhagwatkar

Maîtrise recherche - UdeM

Collaborateur·rice de recherche - UdeM

connor.brennan@mila.quebec

Ethan Caballero

Doctorat - McGill

Superviseur⋅e principal⋅e :

Blake Richards

ethan.caballero@mila.quebec

paolo.cudrano@mila.quebec

Paolo Cudrano

Collaborateur·rice de recherche - Politecnico di Milano

Mohammad-Javad Darvishi Bayazi

Amin Darabi

Doctorat - UdeM

amin.darabi@mila.quebec

Doctorat - UdeM

mohammad-javad.darvishi-bayasi@mila.quebec

Doctorat - UdeM

Co-superviseur⋅e :

Doctorat - UdeM

Collaborateur·rice alumni - UdeM

Co-superviseur⋅e :

Sarath Chandar Anbil Parthipan

kshitij.gupta@mila.quebec

Nadhir Hassen

Stagiaire de recherche - UdeM

nadhir.hassen@mila.quebec

Prateek Humane

Collaborateur·rice de recherche

prateek.humane@mila.quebec

Doctorat - UdeM

Superviseur⋅e principal⋅e :

Nizar Islah

Doctorat - UdeM

Superviseur⋅e principal⋅e :

Eilif Benjamin Muller

nizar.islah@mila.quebec

Doctorat - UdeM

arnav-kumar.jain@mila.quebec

Collaborateur·rice de recherche

ayush.kaushal@mila.quebec

Arian Khorasani

Maîtrise recherche - UdeM

arian.khorasani@mila.quebec

Neeraj Kumar

Maîtrise professionnelle - UdeM

neeraj.kumar@mila.quebec

Gwen Legate

Doctorat - Concordia

Superviseur⋅e principal⋅e :

gwendolyne.legate@mila.quebec

ekaterina.lobacheva@mila.quebec

Kate Lobacheva

Postdoctorat

Superviseur⋅e principal⋅e :

Nicolas Le Roux

Maîtrise recherche - UdeM

sayed.mansouri-tehrani@mila.quebec

juan.mayor-torres@mila.quebec

Juan Manuel Mayor-Torres

Collaborateur·rice de recherche

Visiteur de recherche indépendant

memariaa@mila.quebec

Andrei Mircea Romascanu

Doctorat - UdeM

Maîtrise recherche - UdeM

sparsha.mishra@mila.quebec

Diganta Misra

Maîtrise recherche - UdeM

diganta.misra@mila.quebec

motahareh.pourrahimi@mila.quebec

Remus Mocanu

Maîtrise recherche - UdeM

remus.mocanu@mila.quebec

Timothy Nest

Doctorat - UdeM

Co-superviseur⋅e :

Eilif Benjamin Muller

timothy.nest@mila.quebec

Mohammad Pezeshki

Collaborateur·rice de recherche

Doctorat - McGill

Superviseur⋅e principal⋅e :

Pouya Bashivan

Maximilian Puelma Touzel

Collaborateur·rice de recherche

Maîtrise recherche - UdeM

Co-superviseur⋅e :

mahta.ramezanian@mila.quebec

Roland Riachi

Collaborateur·rice de recherche - UdeM

roland.riachi@mila.quebec

Matthew Riemer

Doctorat - UdeM

matthew.riemer@mila.quebec

Alexis Roger

Maîtrise recherche - UdeM

alexis.roger@mila.quebec

Zahra Sheikhbahaee

Postdoctorat - UdeM

Superviseur⋅e principal⋅e :

zahra.sheikhbahaee@mila.quebec

Vaibhav Singh

Doctorat - Concordia

Superviseur⋅e principal⋅e :

vaibhav.singh@mila.quebec

Gopeshh Subbaraj

Doctorat - UdeM

gopeshh.subbaraj@mila.quebec

Doctorat - UdeM

Co-superviseur⋅e :

benjamin.therien@mila.quebec

Pascal Tikeng Notsawo

Doctorat - UdeM

Co-superviseur⋅e :

pascal.tikeng@mila.quebec

Maîtrise recherche - UdeM

tejas.vaidhya@mila.quebec

Dave Whipps

Maîtrise recherche - UdeM

whippsda@mila.quebec

andrew.williams@mila.quebec

Andrew Williams

Doctorat - UdeM

daria.yasafova@mila.quebec

Daria Yasafova

Stagiaire de recherche - Technical University of Munich

Omar Younis

Stagiaire de recherche - UdeM

omar.younis@mila.quebec

Publications

Challenging Common Assumptions about Catastrophic Forgetting

Timothee LESORT

Oleksiy Ostapenko

Pau Rodriguez

Md Rifat Arefin

Diganta Misra

Laurent Charlin

Building learning agents that can progressively learn and accumulate knowledge is the core goal of the continual learning (CL) research fiel… (voir plus)d. Unfortunately, training a model on new data usually compromises the performance on past data. In the CL literature, this effect is referred to as catastrophic forgetting (CF). CF has been largely studied, and a plethora of methods have been proposed to address it on short sequences of non-overlapping tasks. In such setups, CF always leads to a quick and significant drop in performance in past tasks. Nevertheless, despite CF, recent work showed that SGD training on linear models accumulates knowledge in a CL regression setup. This phenomenon becomes especially visible when tasks reoccur. We might then wonder if DNNs trained with SGD or any standard gradient-based optimization accumulate knowledge in such a way. Such phenomena would have interesting consequences for applying DNNs to real continual scenarios. Indeed, standard gradient-based optimization methods are significantly less computationally expensive than existing CL algorithms. In this paper, we study the progressive knowledge accumulation (KA) in DNNs trained with gradient-based algorithms in long sequences of tasks with data re-occurrence. We propose a new framework, SCoLe (Scaling Continual Learning), to investigate KA and discover that catastrophic forgetting has a limited effect on DNNs trained with SGD. When trained on long sequences with data sparsely re-occurring, the overall accuracy improves, which might be counter-intuitive given the CF phenomenon. We empirically investigate KA in DNNs under various data occurrence frequencies and propose simple and scalable strategies to increase knowledge accumulation in DNNs.

2022-07-10

ArXiv (prépublication)

Towards Scaling Difference Target Propagation by Learning Backprop Targets

Maxence Ernoult

Fabrice Normandin

Abhinav Moudgil

Sean Spinney

The development of biologically-plausible learning algorithms is important for understanding learning in the brain, but most of them fail to… (voir plus) scale-up to real-world tasks, limiting their potential as explanations for learning by real brains. As such, it is important to explore learning algorithms that come with strong theoretical guarantees and can match the performance of backpropagation (BP) on complex tasks. One such algorithm is Difference Target Propagation (DTP), a biologically-plausible learning algorithm whose close relation with Gauss-Newton (GN) optimization has been recently established. However, the conditions under which this connection rigorously holds preclude layer-wise training of the feedback pathway synaptic weights (which is more biologically plausible). Moreover, good alignment between DTP weight updates and loss gradients is only loosely guaranteed and under very specific conditions for the architecture being trained. In this paper, we propose a novel feedback weight training scheme that ensures both that DTP approximates BP and that layer-wise feedback weight training can be restored without sacrificing any theoretical guarantees. Our theory is corroborated by experimental results and we report the best performance ever achieved by DTP on CIFAR-10 and ImageNet 32

2022-06-28

Proceedings of the 39th International Conference on Machine Learning (publié)

proceedings.mlr.press

Parametric Scattering Networks

Shanel Gauthier

Benjamin Thérien

Laurent Alséne-Racicot

Muawiz Chaudhary

Michael Eickenberg

Guy Wolf

The wavelet scattering transform creates geometric in-variants and deformation stability. In multiple signal do-mains, it has been shown to … (voir plus)yield more discriminative rep-resentations compared to other non-learned representations and to outperform learned representations in certain tasks, particularly on limited labeled data and highly structured signals. The wavelet filters used in the scattering trans-form are typically selected to create a tight frame via a pa-rameterized mother wavelet. In this work, we investigate whether this standard wavelet filterbank construction is op-timal. Focusing on Morlet wavelets, we propose to learn the scales, orientations, and aspect ratios of the filters to produce problem-specific parameterizations of the scattering transform. We show that our learned versions of the scattering transform yield significant performance gains in small-sample classification settings over the standard scat-tering transform. Moreover, our empirical results suggest that traditional filterbank constructions may not always be necessary for scattering transforms to extract effective rep-resentations.

2022-06-18

2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (publié)

A Remedy For Distributional Shifts Through Expected Domain Translation

Jean-Christophe Gagnon-Audet

Soroosh Shahtalebi

Frank Rudzicz

Machine learning models often fail to generalize to unseen domains due to the distributional shifts. A family of such shifts, “correlation… (voir plus) shifts,” is caused by spurious correlations in the data. It is studied under the overarching topic of “domain generalization.” In this work, we employ multi-modal translation networks to tackle the correlation shifts that appear when data is sampled out-of-distribution. Learning a generative model from training domains enables us to translate each training sample under the special characteristics of other possible domains. We show that by training a predictor solely on the generated samples, the spurious correlations in training domains average out, and the invariant features corresponding to true correlations emerge. Our proposed technique, Expected Domain Translation (EDT), is benchmarked on the Colored MNIST dataset and drastically improves the state-of-the-art classification accuracy by 38% with train-domain validation model selection.

2022-05-23

IEEE International Conference on Acoustics, Speech, and Signal Processing (publié)

Summarizing Societies: Agent Abstraction in Multi-Agent Reinforcement Learning

Amin Memarian

Maximilian Puelma Touzel

Matthew D Riemer

Rupali Bhati

Agents cannot make sense of many-agent societies through direct consideration of small-scale, low-level agent identities, but instead must r… (voir plus)ecognize emergent collective identities. Here, we take a first step towards a framework for recognizing this structure in large groups of low-level agents so that they can be modeled as a much smaller number of high-level agents—a process that we call agent abstraction. We illustrate this process by extending bisimulation metrics for state abstraction in reinforcement learning to the setting of multi-agent reinforcement learning and analyze a straightforward, if crude, abstraction based on experienced joint actions. It addresses non-stationarity due to other learning agents by improving minimax regret by a intuitive factor. To test if this compression factor provides signal for higher-level agency, we applied it to a large dataset of human play of the popular social dilemma game Diplomacy. We find that it correlates strongly with the degree of ground-truth abstraction of low-level units into the human players.

2022-04-21

ICLR.cc/2022/Workshop/Cells2Societies (poster)

WOODS: Benchmarks for Out-of-Distribution Generalization in Time Series Tasks

Jean-Christophe Gagnon-Audet

Kartik Ahuja

Mohammad-Javad Darvishi-Bayazi

2022-03-18

ArXiv (prépublication)

Cognitive Models as Simulators: The Case of Moral Decision-Making

Ardavan S. Nobandegani

T. Shultz

2022-01-01

CogSci (publié)

Compositional Attention: Disentangling Search and Retrieval

Sarthak Mittal

Sharath Chandra Raparthy

Yoshua Bengio

Guillaume Lajoie

Multi-head, key-value attention is the backbone of transformer-like model architectures which have proven to be widely successful in recent … (voir plus)years. This attention mechanism uses multiple parallel key-value attention blocks (called heads), each performing two fundamental computations: (1) search - selection of a relevant entity from a set via query-key interaction, and (2) retrieval - extraction of relevant features from the selected entity via a value matrix. Standard attention heads learn a rigid mapping between search and retrieval. In this work, we first highlight how this static nature of the pairing can potentially: (a) lead to learning of redundant parameters in certain tasks, and (b) hinder generalization. To alleviate this problem, we propose a novel attention mechanism, called Compositional Attention, that replaces the standard head structure. The proposed mechanism disentangles search and retrieval and composes them in a dynamic, flexible and context-dependent manner. Through a series of numerical experiments, we show that it outperforms standard multi-head attention on a variety of tasks, including some out-of-distribution settings. Through our qualitative analysis, we demonstrate that Compositional Attention leads to dynamic specialization based on the type of retrieval needed. Our proposed mechanism generalizes multi-head attention, allows independent scaling of search and retrieval and is easy to implement in a variety of established network architectures.

2022-01-01

International Conference on Learning Representations (publié)

Continual Learning In Environments With Polynomial Mixing Times

Matthew D Riemer

Sharath Chandra Raparthy

Ignacio Cases

Gopeshh Raaj Subbaraj

Maximilian Puelma Touzel

The mixing time of the Markov chain induced by a policy limits performance in real-world continual learning scenarios. Yet, the effect of mi… (voir plus)xing times on learning in continual reinforcement learning (RL) remains underexplored. In this paper, we characterize problems that are of long-term interest to the development of continual RL, which we call scalable MDPs, through the lens of mixing times. In particular, we theoretically establish that scalable MDPs have mixing times that scale polynomially with the size of the problem. We go on to demonstrate that polynomial mixing times present significant difficulties for existing approaches that suffer from myopic bias and stale bootstrapped estimates. To validate the proposed theory, we study the empirical scaling behavior of mixing times with respect to the number of tasks and task switching frequency for pretrained high performing policies on seven Atari games. Our analysis demonstrates both that polynomial mixing times do emerge in practice and how their existence may lead to unstable learning behavior like catastrophic forgetting in continual learning settings.

Continual Learning with Foundation Models: An Empirical Study of Latent Replay

Oleksiy Ostapenko

Timothee LESORT

Pau Rodriguez

Md Rifat Arefin

Arthur Douillard

Laurent Charlin

2022-01-01

CoLLAs (publié)