Publications

Interference and Generalization in Temporal Difference Learning
Emmanuel Bengio
Joelle Pineau
Doina Precup
We study the link between generalization and interference in temporal-difference (TD) learning. Interference is defined as the inner product… (voir plus) of two different gradients, representing their alignment. This quantity emerges as being of interest from a variety of observations about neural networks, parameter sharing and the dynamics of learning. We find that TD easily leads to low-interference, under-generalizing parameters, while the effect seems reversed in supervised learning. We hypothesize that the cause can be traced back to the interplay between the dynamics of interference and bootstrapping. This is supported empirically by several observations: the negative relationship between the generalization gap and interference in TD, the negative effect of bootstrapping on interference and the local coherence of targets, and the contrast between the propagation rate of information in TD(0) versus TD(
2020-11-20
Proceedings of the 37th International Conference on Machine Learning (publié)
doi.org
proceedings.mlr.press
Invariant Causal Prediction for Block MDPS
Amy Zhang
Clare Lyle
Shagun Sodhani
Angelos Filos
Marta Kwiatkowska
Joelle Pineau
Yarin Gal
Doina Precup
Generalization across environments is critical to the successful application of reinforcement learning algorithms to real-world challenges. … (voir plus)In this paper, we consider the problem of learning abstractions that generalize in block MDPs, families of environments with a shared latent state space and dynamics structure over that latent space, but varying observations. We leverage tools from causal inference to propose a method of invariant prediction to learn model-irrelevance state abstractions (MISA) that generalize to novel observations in the multi-environment setting. We prove that for certain classes of environments, this approach outputs with high probability a state abstraction corresponding to the causal feature set with respect to the return. We further provide more general bounds on model error and generalization error in the multi-environment setting, in the process showing a connection between causal variable selection and the state abstraction framework for MDPs. We give empirical evidence that our methods work in both linear and nonlinear settings, attaining improved generalization over single- and multi-task baselines.
2020-11-20
Proceedings of the 37th International Conference on Machine Learning (publié)
doi.org
proceedings.mlr.press
Latent Variable Modelling with Hyperbolic Normalizing Flows
Avishek Joey Bose
Ariella Smofsky
Renjie Liao
Prakash Panangaden
William L. Hamilton
The choice of approximate posterior distributions plays a central role in stochastic variational inference (SVI). One effective solution is … (voir plus)the use of normalizing flows \cut{defined on Euclidean spaces} to construct flexible posterior distributions. However, one key limitation of existing normalizing flows is that they are restricted to the Euclidean space and are ill-equipped to model data with an underlying hierarchical structure. To address this fundamental limitation, we present the first extension of normalizing flows to hyperbolic spaces. We first elevate normalizing flows to hyperbolic spaces using coupling transforms defined on the tangent bundle, termed Tangent Coupling (
2020-11-20
Proceedings of the 37th International Conference on Machine Learning (publié)
doi.org
proceedings.mlr.press
Linear Lower Bounds and Conditioning of Differentiable Games
Adam Ibrahim
Waïss Azizian
Gauthier Gidel
Ioannis Mitliagkas
Recent successes of game-theoretic formulations in ML have caused a resurgence of research interest in differentiable games. Overwhelmingly,… (voir plus) that research focuses on methods and upper bounds on their speed of convergence. In this work, we approach the question of fundamental iteration complexity by providing lower bounds to complement the linear (i.e. geometric) upper bounds observed in the literature on a wide class of problems. We cast saddle-point and min-max problems as 2-player games. We leverage tools from single-objective convex optimisation to propose new linear lower bounds for convex-concave games. Notably, we give a linear lower bound for
2020-11-20
Proceedings of the 37th International Conference on Machine Learning (publié)
doi.org
proceedings.mlr.press
Perceptual Generative Autoencoders
Zijun Zhang
Ruixiang Zhang
Zongpeng Li
Yoshua Bengio
Liam Paull
Modern generative models are usually designed to match target distributions directly in the data space, where the intrinsic dimension of dat… (voir plus)a can be much lower than the ambient dimension. We argue that this discrepancy may contribute to the difficulties in training generative models. We therefore propose to map both the generated and target distributions to a latent space using the encoder of a standard autoencoder, and train the generator (or decoder) to match the target distribution in the latent space. Specifically, we enforce the consistency in both the data space and the latent space with theoretically justified data and latent reconstruction losses. The resulting generative model, which we call a perceptual generative autoencoder (PGA), is then trained with a maximum likelihood or variational autoencoder (VAE) objective. With maximum likelihood, PGAs generalize the idea of reversible generative models to unrestricted neural network architectures and arbitrary number of latent dimensions. When combined with VAEs, PGAs substantially improve over the baseline VAEs in terms of sample quality. Compared to other autoencoder-based generative models using simple priors, PGAs achieve state-of-the-art FID scores on CIFAR-10 and CelebA.
2020-11-20
Proceedings of the 37th International Conference on Machine Learning (publié)
doi.org
proceedings.mlr.press
Reverse-engineering deep ReLU networks
David Rolnick
Konrad Paul Kording
It has been widely assumed that a neural network cannot be recovered from its outputs, as the network depends on its parameters in a highly … (voir plus)nonlinear way. Here, we prove that in fact it is often possible to identify the architecture, weights, and biases of an unknown deep ReLU network by observing only its output. Every ReLU network defines a piecewise linear function, where the boundaries between linear regions correspond to inputs for which some neuron in the network switches between inactive and active ReLU states. By dissecting the set of region boundaries into components associated with particular neurons, we show both theoretically and empirically that it is possible to recover the weights of neurons and their arrangement within the network, up to isomorphism.
2020-11-20
Proceedings of the 37th International Conference on Machine Learning (publié)
proceedings.mlr.press
proceedings.mlr.press
Revisiting Fundamentals of Experience Replay
William Fedus
Prajit Ramachandran
Rishabh Agarwal
Yoshua Bengio
Hugo Larochelle
Mark Rowland
Will Dabney
Experience replay is central to off-policy algorithms in deep reinforcement learning (RL), but there remain significant gaps in our understa… (voir plus)nding. We therefore present a systematic and extensive analysis of experience replay in Q-learning methods, focusing on two fundamental properties: the replay capacity and the ratio of learning updates to experience collected (replay ratio). Our additive and ablative studies upend conventional wisdom around experience replay -- greater capacity is found to substantially increase the performance of certain algorithms, while leaving others unaffected. Counterintuitively we show that theoretically ungrounded, uncorrected n-step returns are uniquely beneficial while other techniques confer limited benefit for sifting through larger memory. Separately, by directly controlling the replay ratio we contextualize previous observations in the literature and empirically measure its importance across a variety of deep RL algorithms. Finally, we conclude by testing a set of hypotheses on the nature of these performance benefits.
2020-11-20
Proceedings of the 37th International Conference on Machine Learning (publié)
doi.org
proceedings.mlr.press
Universal Equivariant Multilayer Perceptrons
Siamak Ravanbakhsh
Group invariant and equivariant Multilayer Perceptrons (MLP), also known as Equivariant Networks, have achieved remarkable success in learni… (voir plus)ng on a variety of data structures, such as sequences, images, sets, and graphs. Using tools from group theory, this paper proves the universality of a broad class of equivariant MLPs with a single hidden layer. In particular, it is shown that having a hidden layer on which the group acts regularly is sufficient for universal equivariance (invariance). A corollary is unconditional universality of equivariant MLPs for Abelian groups, such as CNNs with a single hidden layer. A second corollary is the universality of equivariant MLPs with a high-order hidden layer, where we give both group-agnostic bounds and means for calculating group-specific bounds on the order of hidden layer that guarantees universal equivariance (invariance).
2020-11-20
Proceedings of the 37th International Conference on Machine Learning (publié)
doi.org
proceedings.mlr.press
What can I do here? A Theory of Affordances in Reinforcement Learning
Khimya Khetarpal
Zafarali Ahmed
Gheorghe Comanici
David Abel
Doina Precup
Reinforcement learning algorithms usually assume that all actions are always available to an agent. However, both people and animals underst… (voir plus)and the general link between the features of their environment and the actions that are feasible. Gibson (1977) coined the term "affordances" to describe the fact that certain states enable an agent to do certain actions, in the context of embodied agents. In this paper, we develop a theory of affordances for agents who learn and plan in Markov Decision Processes. Affordances play a dual role in this case. On one hand, they allow faster planning, by reducing the number of actions available in any given situation. On the other hand, they facilitate more efficient and precise learning of transition models from data, especially when such models require function approximation. We establish these properties through theoretical results as well as illustrative examples. We also propose an approach to learn affordances and use it to estimate transition models that are simpler and generalize better.
2020-11-20
Proceedings of the 37th International Conference on Machine Learning (publié)
doi.org
proceedings.mlr.press
An Effective Anti-Aliasing Approach for Residual Networks
Cristina Vasconcelos
Hugo Larochelle
Vincent Dumoulin
Nicolas Roux
Ross Goroshin
Image pre-processing in the frequency domain has traditionally played a vital role in computer vision and was even part of the standard pipe… (voir plus)line in the early days of deep learning. However, with the advent of large datasets, many practitioners concluded that this was unnecessary due to the belief that these priors can be learned from the data itself. Frequency aliasing is a phenomenon that may occur when sub-sampling any signal, such as an image or feature map, causing distortion in the sub-sampled output. We show that we can mitigate this effect by placing non-trainable blur filters and using smooth activation functions at key locations, particularly where networks lack the capacity to learn them. These simple architectural changes lead to substantial improvements in out-of-distribution generalization on both image classification under natural corruptions on ImageNet-C [10] and few-shot learning on Meta-Dataset [17], without introducing additional trainable parameters and using the default hyper-parameters of open source codebases.
2020-11-19
ArXiv (prépublication)
arxiv.org
arxiv.org
Multiscale PHATE Exploration of SARS-CoV-2 Data Reveals Multimodal Signatures of Disease
Manik Kuchroo
Jessie Huang
Patrick Wong
Jean-Christophe Grenier
Dennis Shung
Alexander Tong
Carolina Lucas
Jon Klein
Daniel B. Burkhardt
Scott Gigante
Abhinav Godavarthi
Benjamin Israelow
Tianyang Mao
Ji Eun Oh
Julio Silva
Takehiro Takahashi
Camila D. Odio
Arnau Casanovas-Massana
John Fournier
Shelli Farhadian … (voir 7 de plus)
Charles S. Dela Cruz
Albert I. Ko
F. Perry Wilson
Julie Hussin
Guy Wolf
Akiko Iwasaki
Smita Krishnaswamy
Abstract

The biomedical community is producing increasingly high dimensional datasets, integrated from hundreds of… (voir plus) patient samples, which current computational techniques struggle to explore. To uncover biological meaning from these complex datasets, we present an approach called Multiscale PHATE, which learns abstracted biological features from data that can be directly predictive of disease. Built on a coarse graining process called diffusion condensation, Multiscale PHATE learns a data topology that can be analyzed at coarse levels for high level summarizations of data, as well as at fine levels for detailed representations on subsets. We apply Multiscale PHATE to study the immune response to COVID-19 in 54 million cells from 168 hospitalized patients. Through our analysis of patient samples, we identify CD16-hi,CD66b-lo neutrophil and IFNγ+,GranzymeB+ Th17 cell responses enriched in patients who die. Furthermore, we show that population groupings Multiscale PHATE discovers can be directly fed into a classifier to predict disease outcome. We also use Multiscale PHATE-derived features to construct two different manifolds of patients, one from abstracted flow cytometry features and another directly on patient clinical features, both associating immune subsets and clinical markers with outcome.

2020-11-16
bioRxiv (prépublication)
doi.org
Using Open Source Licensing to Regulate the Assembly of LAWS: A Preliminary Analysis
Cheng Lin
AJung Moon
Lethal autonomous weapons (LAWS) are an emerging technology capable of automatically targeting and exercising lethal force. Many scholars an… (voir plus)d advocates have petitioned to ban the technology internationally for a myriad of reasons. However, there are practical challenges to implementing a ban. One such challenge is posed by the “intangible” nature of the software that LAWS depends on, which is incompatible with implementation mechanisms such as export control. Given the dual-use nature of software, and the fact that software is developed by teams of individuals, a number of soft governance mechanisms have been proposed to regulate this technology. In this paper, we investigate the feasibility of one particular approach: leveraging open source licenses as a means to prohibit the use of certain software in LAWS. This approach is largely motivated by the fact that open source software underpins all of technology, especially AI. Through a review of the recent tech activism and open source activism, we evaluate whether open source licenses can feasibly limit the use of open source software to only non-LAWS applications. We distill the current challenges facing “ethics-driven” open source licensing efforts into three main obstacles: the need for clarity of licensing language, the lack of enforceability of licenses, and the lack of cohesiveness of the open source community. We propose that addressing these factors are also success criteria for future anti-LAWS open source initiatives. We find that open source licenses provide more theoretical than practical promise in regulating LAWS, and conclude that cohesion in the open source community is the key to their potential practical success in the future.
2020-11-11
2020 IEEE International Symposium on Technology and Society (ISTAS) (publié)
doi.org