date,title,abstract,keywords,type,accepted
28 Sept 2020,What Matters for On-Policy Deep Actor-Critic Methods? A Large-Scale Study ,"In recent years, reinforcement learning (RL) has been successfully applied to many different continuous control tasks. While RL algorithms are often conceptually simple, their state-of-the-art implementations take numerous low- and high-level design decisions that strongly affect the performance of the resulting agents. Those choices are usually not extensively discussed in the literature, leading to discrepancy between published descriptions of algorithms and their implementations. This makes it hard to attribute progress in RL and slows down overall progress [Engstrom'20]. As a step towards filling that gap, we implement >50 such ``""choices"" in a unified on-policy deep actor-critic framework, allowing us to investigate their impact in a large-scale empirical study. We train over 250'000 agents in five continuous control environments of different complexity and provide insights and practical recommendations for the training of on-policy deep actor-critic RL agents.","Reinforcement learning, continuous control",oral presentations,1
28 Sept 2020,Theoretical Analysis of Self-Training with Deep Networks on Unlabeled Data,"Self-training algorithms, which train a model to fit pseudolabels predicted by another previously-learned model, have been very successful for learning with unlabeled data using neural networks. However, the current theoretical understanding of self-training only applies to linear models. This work provides a unified theoretical analysis of self-training with deep networks for semi-supervised learning, unsupervised domain adaptation, and unsupervised learning. At the core of our analysis is a simple but realistic “expansion” assumption, which states that a low-probability subset of the data must expand to a neighborhood with large probability relative to the subset. We also assume that neighborhoods of examples in different classes have minimal overlap. We prove that under these assumptions, the minimizers of population objectives based on self-training and input-consistency regularization will achieve high accuracy with respect to ground-truth labels. By using off-the-shelf generalization bounds, we immediately convert this result to sample complexity guarantees for neural nets that are polynomial in the margin and Lipschitzness. Our results help explain the empirical successes of recently proposed self-training algorithms which use input consistency regularization.","deep learning theory, domain adaptation theory, unsupervised learning theory, semi-supervised learning theory",oral presentations,1
28 Sept 2020,Learning to Reach Goals via Iterated Supervised Learning,"Current reinforcement learning (RL) algorithms can be brittle and difficult to use, especially when learning goal-reaching behaviors from sparse rewards. Although supervised imitation learning provides a simple and stable alternative, it requires access to demonstrations from a human supervisor. In this paper, we study RL algorithms that use imitation learning to acquire goal reaching policies from scratch, without the need for expert demonstrations or a value function. In lieu of demonstrations, we leverage the property that any trajectory is a successful demonstration for reaching the final state in that same trajectory. We propose a simple algorithm in which an agent continually relabels and imitates the trajectories it generates to progressively learn goal-reaching behaviors from scratch. Each iteration, the agent collects new trajectories using the latest policy, and maximizes the likelihood of the actions along these trajectories under the goal that was actually reached, so as to improve the policy. We formally show that this iterated supervised learning procedure optimizes a bound on the RL objective, derive performance bounds of the learned policy, and empirically demonstrate improved goal-reaching performance and robustness over current RL algorithms in several benchmark tasks. ","goal reaching, reinforcement learning, behavior cloning, goal-conditioned RL",oral presentations,1
28 Sept 2020,Deep symbolic regression: Recovering mathematical expressions from data via risk-seeking policy gradients,"Discovering the underlying mathematical expressions describing a dataset is a core challenge for artificial intelligence. This is the problem of . Despite recent advances in training neural networks to solve complex tasks, deep learning approaches to symbolic regression are underexplored. We propose a framework that leverages deep learning for symbolic regression via a simple idea: use a large model to search the space of small models. Specifically, we use a recurrent neural network to emit a distribution over tractable mathematical expressions and employ a novel risk-seeking policy gradient to train the network to generate better-fitting expressions. Our algorithm outperforms several baseline methods (including Eureqa, the gold standard for symbolic regression) in its ability to exactly recover symbolic expressions on a series of benchmark problems, both with and without added noise. More broadly, our contributions include a framework that can be applied to optimize hierarchical, variable-length objects under a black-box performance metric, with the ability to incorporate constraints in situ, and a risk-seeking policy gradient formulation that optimizes for best-case performance instead of expected performance.","symbolic regression, reinforcement learning, automated machine learning",oral presentations,1
28 Sept 2020,Optimal Rates for Averaged Stochastic Gradient Descent under Neural Tangent Kernel Regime ,"We analyze the convergence of the averaged stochastic gradient descent for overparameterized two-layer neural networks for regression problems. It was recently found that a neural tangent kernel (NTK) plays an important role in showing the global convergence of gradient-based methods under the NTK regime, where the learning dynamics for overparameterized neural networks can be almost characterized by that for the associated reproducing kernel Hilbert space (RKHS). However, there is still room for a convergence rate analysis in the NTK regime. In this study, we show that the averaged stochastic gradient descent can achieve the minimax optimal convergence rate, with the global convergence guarantee, by exploiting the complexities of the target function and the RKHS associated with the NTK. Moreover, we show that the target function specified by the NTK of a ReLU network can be learned at the optimal convergence rate through a smooth approximation of a ReLU network under certain conditions.","stochastic gradient descent, two-layer neural network, over-parameterization, neural tangent kernel",oral presentations,1
28 Sept 2020,Free Lunch for Few-shot Learning: Distribution Calibration,"Learning from a limited number of samples is challenging since the learned model can easily become overfitted based on the biased distribution formed by only a few training examples. In this paper, we calibrate the distribution of these few-sample classes by transferring statistics from the classes with sufficient examples. Then an adequate number of examples can be sampled from the calibrated distribution to expand the inputs to the classifier. We assume every dimension in the feature representation follows a Gaussian distribution so that the mean and the variance of the distribution can borrow from that of similar classes whose statistics are better estimated with an adequate number of samples. Our method can be built on top of off-the-shelf pretrained feature extractors and classification models without extra parameters. We show that a simple logistic regression classifier trained using the features sampled from our calibrated distribution can outperform the state-of-the-art accuracy on three datasets (~5% improvement on miniImageNet compared to the next best). The visualization of these generated features demonstrates that our calibrated distribution is an accurate estimation. ","few-shot learning, image classification, distribution estimation",oral presentations,1
28 Sept 2020,Scalable Learning and MAP Inference for Nonsymmetric Determinantal Point Processes,"Determinantal point processes (DPPs) have attracted significant attention in machine learning for their ability to model subsets drawn from a large item collection. Recent work shows that nonsymmetric DPP (NDPP) kernels have significant advantages over symmetric kernels in terms of modeling power and predictive performance. However, for an item collection of size , existing NDPP learning and inference algorithms require memory quadratic in  and runtime cubic (for learning) or quadratic (for inference) in , making them impractical for many typical subset selection tasks. In this work, we develop a learning algorithm with space and time requirements linear in  by introducing a new NDPP kernel decomposition. We also derive a linear-complexity NDPP maximum a posteriori (MAP) inference algorithm that applies not only to our new kernel but also to that of prior work. Through evaluation on real-world datasets, we show that our algorithms scale significantly better, and can match the predictive performance of prior work.","determinantal point processes, unsupervised learning, representation learning, submodular optimization",oral presentations,1
28 Sept 2020,Randomized Automatic Differentiation,"The successes of deep learning, variational inference, and many other fields have been aided by specialized implementations of reverse-mode automatic differentiation (AD) to compute gradients of mega-dimensional objectives. The AD techniques underlying these tools were designed to compute exact gradients to numerical precision, but modern machine learning models are almost always trained with stochastic gradient descent. Why spend computation and memory on exact (minibatch) gradients only to use them for stochastic optimization? We develop a general framework and approach for randomized automatic differentiation (RAD), which can allow unbiased gradient estimates to be computed with reduced memory in return for variance. We examine limitations of the general approach, and argue that we must leverage problem specific structure to realize benefits. We develop RAD techniques for a variety of simple neural network architectures, and show that for a fixed memory budget, RAD converges in fewer iterations than using a small batch size for feedforward networks, and in a similar number for recurrent networks. We also show that RAD can be applied to scientific computing, and use it to develop a low-memory stochastic gradient method for optimizing the control parameters of a linear reaction-diffusion PDE representing a fission reactor.","automatic differentiation, autodiff, backprop, deep learning, pdes, stochastic optimization",oral presentations,1
28 Sept 2020,Learning Generalizable Visual Representations via Interactive Gameplay,"A growing body of research suggests that embodied gameplay, prevalent not just in human cultures but across a variety of animal species including turtles and ravens, is critical in developing the neural flexibility for creative problem solving, decision making, and socialization. Comparatively little is known regarding the impact of embodied gameplay upon artificial agents. While recent work has produced agents proficient in abstract games, these environments are far removed the real world and thus these agents can provide little insight into the advantages of embodied play. Hiding games, such as hide-and-seek, played universally, provide a rich ground for studying the impact of embodied gameplay on representation learning in the context of perspective taking, secret keeping, and false belief understanding. Here we are the first to show that embodied adversarial reinforcement learning agents playing Cache, a variant of hide-and-seek, in a high fidelity, interactive, environment, learn generalizable representations of their observations encoding information such as object permanence, free space, and containment. Moving closer to biologically motivated learning strategies, our agents' representations, enhanced by intentionality and memory, are developed through interaction and play. These results serve as a model for studying how facets of vision develop through interaction, provide an experimental framework for assessing what is learned by artificial agents, and demonstrates the value of moving from large, static, datasets towards experiential, interactive, representation learning.","representation learning, deep reinforcement learning, computer vision",oral presentations,1
28 Sept 2020,Global Convergence of Three-layer Neural Networks in the Mean Field Regime,"In the mean field regime, neural networks are appropriately scaled so that as the width tends to infinity, the learning dynamics tends to a nonlinear and nontrivial dynamical limit, known as the mean field limit. This lends a way to study large-width neural networks via analyzing the mean field limit. Recent works have successfully applied such analysis to two-layer networks and provided global convergence guarantees. The extension to multilayer ones however has been a highly challenging puzzle, and little is known about the optimization efficiency in the mean field regime when there are more than two layers. In this work, we prove a global convergence result for unregularized feedforward three-layer networks in the mean field regime. We first develop a rigorous framework to establish the mean field limit of three-layer networks under stochastic gradient descent training. To that end, we propose the idea of a neuronal embedding, which comprises of a fixed probability space that encapsulates neural networks of arbitrary sizes. The identified mean field limit is then used to prove a global convergence guarantee under suitable regularity and convergence mode assumptions, which – unlike previous works on two-layer networks – does not rely critically on convexity. Underlying the result is a universal approximation property, natural of neural networks, which importantly is shown to hold at any finite training time (not necessarily at convergence) via an algebraic topology argument.",deep learning theory,oral presentations,1
28 Sept 2020,Rao-Blackwellizing the Straight-Through Gumbel-Softmax Gradient Estimator ,"Gradient estimation in models with discrete latent variables is a challenging problem, because the simplest unbiased estimators tend to have high variance. To counteract this, modern estimators either introduce bias, rely on multiple function evaluations, or use learned, input-dependent baselines. Thus, there is a need for estimators that require minimal tuning, are computationally cheap, and have low mean squared error. In this paper, we show that the variance of the straight-through variant of the popular Gumbel-Softmax estimator can be reduced through Rao-Blackwellization without increasing the number of function evaluations. This provably reduces the mean squared error. We empirically demonstrate that this leads to variance reduction, faster convergence, and generally improved performance in two unsupervised latent variable models.","gumbel, softmax, gumbel-softmax, straight-through, straightthrough, rao, rao-blackwell",oral presentations,1
28 Sept 2020,Rethinking Attention with Performers,"We introduce Performers, Transformer architectures which can estimate regular (softmax) full-rank-attention Transformers with provable accuracy, but using only linear (as opposed to quadratic) space and time complexity, without relying on any priors such as sparsity or low-rankness. To approximate softmax attention-kernels, Performers use a novel Fast Attention Via positive Orthogonal Random features approach (FAVOR+), which may be of independent interest for scalable kernel methods. FAVOR+ can also be used to efficiently model kernelizable attention mechanisms beyond softmax. This representational power is crucial to accurately compare softmax with other kernels for the first time on large-scale tasks, beyond the reach of regular Transformers, and investigate optimal attention-kernels. Performers are linear architectures fully compatible with regular Transformers and with strong theoretical guarantees: unbiased or nearly-unbiased estimation of the attention matrix, uniform convergence and low  estimation variance. We tested Performers on a rich set of tasks stretching from pixel-prediction through text models to protein sequence modeling. We demonstrate competitive results with other examined efficient sparse and dense attention methods, showcasing effectiveness of the novel attention-learning paradigm leveraged by Performers. ","performer, transformer, attention, softmax, approximation, linear, bert, bidirectional, unidirectional, orthogonal, random, features, FAVOR, kernel, generalized, sparsity, reformer, linformer, protein, trembl, uniprot",oral presentations,1
28 Sept 2020,Getting a CLUE: A Method for Explaining Uncertainty Estimates,"Both uncertainty estimation and interpretability are important factors for trustworthy machine learning systems. However, there is little work at the intersection of these two areas. We address this gap by proposing a novel method for interpreting uncertainty estimates from differentiable probabilistic models, like Bayesian Neural Networks (BNNs). Our method, Counterfactual Latent Uncertainty Explanations (CLUE), indicates how to change an input, while keeping it on the data manifold, such that a BNN becomes more confident about the input's prediction. We validate CLUE through 1) a novel framework for evaluating counterfactual explanations of uncertainty, 2) a series of ablation experiments, and 3) a user study. Our experiments show that CLUE outperforms baselines and enables practitioners to better understand which input patterns are responsible for predictive uncertainty.","interpretability, uncertainty, explainability",oral presentations,1
28 Sept 2020,When Do Curricula Work? ,"Inspired by human learning, researchers have proposed ordering examples during training based on their difficulty. Both curriculum learning, exposing a network to easier examples early in training, and anti-curriculum learning, showing the most difficult examples first, have been suggested as improvements to the standard i.i.d. training. In this work, we set out to investigate the relative benefits of ordered learning. We first investigate the implicit curricula resulting from architectural and optimization bias and find that samples are learned in a highly consistent order. Next, to quantify the benefit of explicit curricula, we conduct extensive experiments over thousands of orderings spanning three kinds of learning: curriculum, anti-curriculum, and random-curriculum -- in which the size of the training dataset is dynamically increased over time, but the examples are randomly ordered. We find that for standard benchmark datasets, curricula have only marginal benefits, and that randomly ordered samples perform as well or better than curricula and anti-curricula, suggesting that any benefit is entirely due to the dynamic training set size. Inspired by common use cases of curriculum learning in practice, we investigate the role of limited training time budget and noisy data in the success of curriculum learning. Our experiments demonstrate that curriculum, but not anti-curriculum or random ordering can indeed improve the performance either with limited training time budget or in the existence of noisy data.","Curriculum Learning, Understanding Deep Learning, Empirical Investigation",oral presentations,1
28 Sept 2020,Federated Learning Based on Dynamic Regularization,"We propose a novel federated learning method for distributively training neural network models, where the server orchestrates cooperation between a subset of randomly chosen devices in each round. We view Federated Learning problem primarily from a communication perspective and allow more device level computations to save transmission costs. We point out a fundamental dilemma, in that the minima of the local-device level empirical loss are inconsistent with those of the global empirical loss. Different from recent prior works, that either attempt inexact minimization or utilize devices for parallelizing gradient computation, we propose a dynamic regularizer for each device at each round, so that in the limit the global and device solutions are aligned. We demonstrate both through empirical results on real and synthetic data as well as analytical results that our scheme leads to efficient training, in both convex and non-convex settings, while being fully agnostic to device heterogeneity and robust to large number of devices, partial participation and unbalanced data.","Federated Learning, Deep Neural Networks, Distributed Optimization",oral presentations,1
28 Sept 2020,Geometry-aware Instance-reweighted Adversarial Training,"In adversarial machine learning, there was a common belief that robustness and accuracy hurt each other. The belief was challenged by recent studies where we can maintain the robustness and improve the accuracy. However, the other direction, whether we can keep the accuracy and improve the robustness, is conceptually and practically more interesting, since robust accuracy should be lower than standard accuracy for any model. In this paper, we show this direction is also promising. Firstly, we find even over-parameterized deep networks may still have insufficient model capacity, because adversarial training has an overwhelming smoothing effect. Secondly, given limited model capacity, we argue adversarial data should have unequal importance: geometrically speaking, a natural data point closer to/farther from the class boundary is less/more robust, and the corresponding adversarial data point should be assigned with larger/smaller weight. Finally, to implement the idea, we propose geometry-aware instance-reweighted adversarial training, where the weights are based on how difficult it is to attack a natural data point. Experiments show that our proposal boosts the robustness of standard adversarial training; combining two directions, we improve both robustness and accuracy of standard adversarial training.",Adversarial robustness,oral presentations,1
28 Sept 2020,Co-Mixup: Saliency Guided Joint Mixup with Supermodular Diversity ,"While deep neural networks show great performance on fitting to the training distribution, improving the networks' generalization performance to the test distribution and robustness to the sensitivity to input perturbations still remain as a challenge. Although a number of mixup based augmentation strategies have been proposed to partially address them, it remains unclear as to how to best utilize the supervisory signal within each input data for mixup from the optimization perspective. We propose a new perspective on batch mixup and formulate the optimal construction of a batch of mixup data maximizing the data saliency measure of each individual mixup data and encouraging the supermodular diversity among the constructed mixup data. This leads to a novel discrete optimization problem minimizing the difference between submodular functions. We also propose an efficient modular approximation based iterative submodular minimization algorithm for efficient mixup computation per each minibatch suitable for minibatch based neural network training. Our experiments show the proposed method achieves the state of the art generalization, calibration, and weakly supervised localization results compared to other mixup methods. The source code is available at https://github.com/snu-mllab/Co-Mixup.","Data Augmentation, Deep Learning, Supervised Learning, Discrete Optimization",oral presentations,1
28 Sept 2020,SenSeI: Sensitive Set Invariance for Enforcing Individual Fairness,"In this paper, we cast fair machine learning as invariant machine learning. We first formulate a version of individual fairness that enforces invariance on certain sensitive sets. We then design a transport-based regularizer that enforces this version of individual fairness and develop an algorithm to minimize the regularizer efficiently. Our theoretical results guarantee the proposed approach trains certifiably fair ML models. Finally, in the experimental studies we demonstrate improved fairness metrics in comparison to several recent fair training procedures on three ML tasks that are susceptible to algorithmic bias.","Algorithmic fairness, invariance",oral presentations,1
28 Sept 2020,End-to-end Adversarial Text-to-Speech ,"Modern text-to-speech synthesis pipelines typically involve multiple processing stages, each of which is designed or learnt independently from the rest. In this work, we take on the challenging task of learning to synthesise speech from normalised text or phonemes in an end-to-end manner, resulting in models which operate directly on character or phoneme input sequences and produce raw speech audio outputs. Our proposed generator is feed-forward and thus efficient for both training and inference, using a differentiable alignment scheme based on token length prediction. It learns to produce high fidelity audio through a combination of adversarial feedback and prediction losses constraining the generated audio to roughly match the ground truth in terms of its total duration and mel-spectrogram. To allow the model to capture temporal variation in the generated audio, we employ soft dynamic time warping in the spectrogram-based prediction loss. The resulting model achieves a mean opinion score exceeding 4 on a 5 point scale, which is comparable to the state-of-the-art models relying on multi-stage training and additional supervision.","text-to-speech, speech synthesis, adversarial, GAN, end-to-end, feed-forward, generative model",oral presentations,1
28 Sept 2020,Dataset Condensation with Gradient Matching,"As the state-of-the-art machine learning methods in many fields rely on larger datasets, storing datasets and training models on them become significantly more expensive. This paper proposes a training set synthesis technique for data-efficient learning, called Dataset Condensation, that learns to condense large dataset into a small set of informative synthetic samples for training deep neural networks from scratch. We formulate this goal as a gradient matching problem between the gradients of deep neural network weights that are trained on the original and our synthetic data. We rigorously evaluate its performance in several computer vision benchmarks and demonstrate that it significantly outperforms the state-of-the-art methods. Finally we explore the use of our method in continual learning and neural architecture search and report promising gains when limited memory and computations are available.","dataset condensation, data-efficient learning, image generation",oral presentations,1
28 Sept 2020,Rethinking Architecture Selection in Differentiable NAS ,"Differentiable Neural Architecture Search is one of the most popular Neural Architecture Search (NAS) methods for its search efficiency and simplicity, accomplished by jointly optimizing the model weight and architecture parameters in a weight-sharing supernet via gradient-based algorithms. At the end of the search phase, the operations with the largest architecture parameters will be selected to form the final architecture, with the implicit assumption that the values of architecture parameters reflect the operation strength. While much has been discussed about the supernet's optimization, the architecture selection process has received little attention. We provide empirical and theoretical analysis to show that the magnitude of architecture parameters does not necessarily indicate how much the operation contributes to the supernet's performance. We propose an alternative perturbation-based architecture selection that directly measures each operation's influence on the supernet. We re-evaluate several differentiable NAS methods with the proposed architecture selection and find that it is able to extract significantly improved architectures from the underlying supernets consistently. Furthermore, we find that several failure modes of DARTS can be greatly alleviated with the proposed selection method, indicating that much of the poor generalization observed in DARTS can be attributed to the failure of magnitude-based architecture selection rather than entirely the optimization of its supernet.",,oral presentations,1
28 Sept 2020,A Distributional Approach to Controlled Text Generation ,"We propose a  Distributional  Approach for addressing  Controlled  Text  Generation from pre-trained Language Models (LM). This approach permits to specify, in a single formal framework, both “pointwise’” and “distributional” constraints over the target LM — to our knowledge, the first model with such generality —while minimizing KL divergence from the initial LM distribution.  The optimal target distribution is then uniquely determined as an explicit EBM (Energy-BasedModel) representation. From that optimal representation, we then train a target controlled Autoregressive LM through an adaptive distributional variant of PolicyGradient.  We conduct a first set of experiments over pointwise constraints showing the advantages of our approach over a set of baselines, in terms of obtaining a controlled LM balancing constraint satisfaction with divergence from the pretrained LM.  We then perform experiments over distributional constraints, a unique feature of our approach, demonstrating its potential as a remedy to the problem of Bias in Language Models.  Through an ablation study, we show the effectiveness of our adaptive technique for obtaining faster convergence.","Controlled NLG, Pretrained Language Models, Bias in Language Models, Energy-Based Models, Information Geometry, Exponential Families",oral presentations,1
28 Sept 2020,Learning Cross-Domain Correspondence for Control with Dynamics Cycle-Consistency ,"At the heart of many robotics problems is the challenge of learning correspondences across domains. For instance, imitation learning requires obtaining correspondence between humans and robots; sim-to-real requires correspondence between physics simulators and real hardware; transfer learning requires correspondences between different robot environments. In this paper, we propose to learn correspondence across such domains emphasizing on differing modalities (vision and internal state), physics parameters (mass and friction), and morphologies (number of limbs). Importantly, correspondences are learned using unpaired and randomly collected data from the two domains. We propose dynamics cycles that align dynamic robotic behavior across two domains using a cycle consistency constraint. Once this correspondence is found, we can directly transfer the policy trained on one domain to the other, without needing any additional fine-tuning on the second domain. We perform experiments across a variety of problem domains, both in simulation and on real robots. Our framework is able to align uncalibrated monocular video of a real robot arm to dynamic state-action trajectories of a simulated arm without paired data. Video demonstrations of our results are available at: https://sites.google.com/view/cycledynamics .","self-supervised learning, robotics",oral presentations,1
28 Sept 2020,Human-Level Performance in No-Press Diplomacy via Equilibrium Search ,"Prior AI breakthroughs in complex games have focused on either the purely adversarial or purely cooperative settings. In contrast, Diplomacy is a game of shifting alliances that involves both cooperation and competition. For this reason, Diplomacy has proven to be a formidable research challenge. In this paper we describe an agent for the no-press variant of Diplomacy that combines supervised learning on human data with one-step lookahead search via regret minimization. Regret minimization techniques have been behind previous AI successes in adversarial games, most notably poker, but have not previously been shown to be successful in large-scale games involving cooperation. We show that our agent greatly exceeds the performance of past no-press Diplomacy bots, is unexploitable by expert humans, and ranks in the top 2% of human players when playing anonymous games on a popular Diplomacy website.","multi-agent systems, regret minimization, no-regret learning, game theory, reinforcement learning",oral presentations,1
28 Sept 2020,Parrot: Data-Driven Behavioral Priors for Reinforcement Learning ,"Reinforcement learning provides a general framework for flexible decision making and control, but requires extensive data collection for each new task that an agent needs to learn. In other machine learning fields, such as natural language processing or computer vision, pre-training on large, previously collected datasets to bootstrap learning for new tasks has emerged as a powerful paradigm to reduce data requirements when learning a new task. In this paper, we ask the following question: how can we enable similarly useful pre-training for RL agents? We propose a method for pre-training behavioral priors that can capture complex input-output relationships observed in successful trials from a wide range of previously seen tasks, and we show how this learned prior can be used for rapidly learning new tasks without impeding the RL agent's ability to try out novel behaviors. We demonstrate the effectiveness of our approach in challenging robotic manipulation domains involving image observations and sparse reward functions, where our method outperforms prior works by a substantial margin. Additional materials can be found on our project website: https://sites.google.com/view/parrot-rl","reinforcement learning, imitation learning",oral presentations,1
28 Sept 2020,Learning Invariant Representations for Reinforcement Learning without Reconstruction ,"We study how representation learning can accelerate reinforcement learning from rich observations, such as images, without relying either on domain knowledge or pixel-reconstruction. Our goal is to learn representations that provide for effective downstream control and invariance to task-irrelevant details. Bisimulation metrics quantify behavioral similarity between states in continuous MDPs, which we propose using to learn robust latent representations which encode only the task-relevant information from observations. Our method trains encoders such that distances in latent space equal bisimulation distances in state space. We demonstrate the effectiveness of our method at disregarding task-irrelevant information using modified visual MuJoCo tasks, where the background is replaced with moving distractors and natural videos, while achieving SOTA performance. We also test a first-person highway driving task where our method learns invariance to clouds, weather, and time of day. Finally, we provide generalization results drawn from properties of bisimulation metrics, and links to causal inference.","rich observations, bisimulation metrics, representation learning, state abstractions",oral presentations,1
28 Sept 2020,Do 2D GANs Know 3D Shape? Unsupervised 3D Shape Reconstruction from 2D Image GANs ,"Natural images are projections of 3D objects on a 2D image plane. While state-of-the-art 2D generative models like GANs show unprecedented quality in modeling the natural image manifold, it is unclear whether they implicitly capture the underlying 3D object structures. And if so, how could we exploit such knowledge to recover the 3D shapes of objects in the images? To answer these questions, in this work, we present the first attempt to directly mine 3D geometric cues from an off-the-shelf 2D GAN that is trained on RGB images only. Through our investigation, we found that such a pre-trained GAN indeed contains rich 3D knowledge and thus can be used to recover 3D shape from a single 2D image in an unsupervised manner. The core of our framework is an iterative strategy that explores and exploits diverse viewpoint and lighting variations in the GAN image manifold. The framework does not require 2D keypoint or 3D annotations, or strong assumptions on object shapes (e.g. shapes are symmetric), yet it successfully recovers 3D shapes with high precision for human faces, cats, cars, and buildings. The recovered 3D shapes immediately allow high-quality image editing like relighting and object rotation. We quantitatively demonstrate the effectiveness of our approach compared to previous methods in both 3D shape reconstruction and face rotation. Our code is available at https://github.com/XingangPan/GAN2Shape.","Generative Adversarial Network, 3D Reconstruction",oral presentations,1
28 Sept 2020,VCNet and Functional Targeted Regularization For Learning Causal Effects of Continuous Treatments ,"Motivated by the rising abundance of observational data with continuous treatments, we investigate the problem of estimating the average dose-response curve (ADRF). Available parametric methods are limited in their model space, and previous attempts in leveraging neural network to enhance model expressiveness relied on partitioning continuous treatment into blocks and using separate heads for each block; this however produces in practice discontinuous ADRFs. Therefore, the question of how to adapt the structure and training of neural network to estimate ADRFs remains open. This paper makes two important contributions. First, we propose a novel varying coefficient neural network (VCNet) that improves model expressiveness while preserving continuity of the estimated ADRF. Second, to improve finite sample performance, we generalize targeted regularization to obtain a doubly robust estimator of the whole ADRF curve.","causal inference, continuous treatment effect, doubly robustness",oral presentations,1
28 Sept 2020,Rethinking the Role of Gradient-based Attribution Methods for Model Interpretability ,"Current methods for the interpretability of discriminative deep neural networks commonly rely on the model's input-gradients, i.e., the gradients of the output logits w.r.t. the inputs. The common assumption is that these input-gradients contain information regarding , the model's discriminative capabilities, thus justifying their use for interpretability. However, in this work, we show that these input-gradients can be arbitrarily manipulated as a consequence of the shift-invariance of softmax without changing the discriminative function. This leaves an open question: given that input-gradients can be arbitrary, why are they highly structured and explanatory in standard models? In this work, we re-interpret the logits of standard softmax-based classifiers as unnormalized log-densities of the data distribution and show that input-gradients can be viewed as gradients of a class-conditional generative model  implicit in the discriminative model. This leads us to hypothesize that the highly structured and explanatory nature of input-gradients may be due to the alignment of this class-conditional model  with that of the ground truth data distribution . We test this hypothesis by studying the effect of density alignment on gradient explanations. To achieve this density alignment, we use an algorithm called score-matching, and propose novel approximations to this algorithm to enable training large-scale models.
 Our experiments show that improving the alignment of the implicit density model with the data distribution enhances gradient structure and explanatory power while reducing this alignment has the opposite effect. This also leads us to conjecture that unintended density alignment in standard neural network training may explain the highly structured nature of input-gradients observed in practice. Overall, our finding that input-gradients capture information regarding an implicit generative model implies that we need to re-think their use for interpreting discriminative models.","Interpretability, saliency maps, score-matching",oral presentations,1
28 Sept 2020,Neural Synthesis of Binaural Speech From Mono Audio ,"We present a neural rendering approach for binaural sound synthesis that can produce realistic and spatially accurate binaural sound in realtime. The network takes, as input, a single-channel audio source and synthesizes, as output, two-channel binaural sound, conditioned on the relative position and orientation of the listener with respect to the source. We investigate deficiencies of the l2-loss on raw waveforms in a theoretical analysis and introduce an improved loss that overcomes these limitations. In an empirical evaluation, we establish that our approach is the first to generate spatially accurate waveform outputs (as measured by real recordings) and outperforms existing approaches by a considerable margin, both quantitatively and in a perceptual study. Dataset and code are available online.","binaural audio, sound spatialization, neural sound synthesis, binaural speech, speech processing, speech generation",oral presentations,1
28 Sept 2020,DiffWave: A Versatile Diffusion Model for Audio Synthesis ,"In this work, we propose DiffWave, a versatile diffusion probabilistic model for conditional and unconditional waveform generation. The model is non-autoregressive, and converts the white noise signal into structured waveform through a Markov chain with a constant number of steps at synthesis. It is efficiently trained by optimizing a variant of variational bound on the data likelihood. DiffWave produces high-fidelity audios in different waveform generation tasks, including neural vocoding conditioned on mel spectrogram, class-conditional generation, and unconditional generation. We demonstrate that DiffWave matches a strong WaveNet vocoder in terms of speech quality (MOS: 4.44 versus 4.43), while synthesizing orders of magnitude faster. In particular, it significantly outperforms autoregressive and GAN-based waveform models in the challenging unconditional generation task in terms of audio quality and sample diversity from various automatic and human evaluations.","diffusion probabilistic models, audio synthesis, speech synthesis, generative models",oral presentations,1
28 Sept 2020,An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale ,"While the Transformer architecture has become the de-facto standard for natural language processing tasks, its applications to computer vision remain limited. In vision, attention is either applied in conjunction with convolutional networks, or used to replace certain components of convolutional networks while keeping their overall structure in place. We show that this reliance on CNNs is not necessary and a pure transformer applied directly to sequences of image patches can perform very well on image classification tasks. When pre-trained on large amounts of data and transferred to multiple mid-sized or small image recognition benchmarks (ImageNet, CIFAR-100, VTAB, etc.), Vision Transformer (ViT) attains excellent results compared to state-of-the-art convolutional networks while requiring substantially fewer computational resources to train.","computer vision, image recognition, self-attention, transformer, large-scale training",oral presentations,1
28 Sept 2020,On the mapping between Hopfield networks and Restricted Boltzmann Machines,"Hopfield networks (HNs) and Restricted Boltzmann Machines (RBMs) are two important models at the interface of statistical physics, machine learning, and neuroscience. Recently, there has been interest in the relationship between HNs and RBMs, due to their similarity under the statistical mechanics formalism. An exact mapping between HNs and RBMs has been previously noted for the special case of orthogonal (“uncorrelated”) encoded patterns. We present here an exact mapping in the case of correlated pattern HNs, which are more broadly applicable to existing datasets. Specifically, we show that any HN with  binary variables and  potentially correlated binary patterns can be transformed into an RBM with  binary visible variables and  gaussian hidden variables. We outline the conditions under which the reverse mapping exists, and conduct experiments on the MNIST dataset which suggest the mapping provides a useful initialization to the RBM weights. We discuss extensions, the potential importance of this correspondence for the training of RBMs, and for understanding the performance of feature extraction methods which utilize RBMs.","Hopfield Networks, Restricted Boltzmann Machines, Statistical Physics",oral presentations,1
28 Sept 2020,SMiRL: Surprise Minimizing Reinforcement Learning in Unstable Environments,"Every living organism struggles against disruptive environmental forces to carve out and maintain an orderly niche. We propose that such a struggle to achieve and preserve order might offer a principle for the emergence of useful behaviors in artificial agents. We formalize this idea into an unsupervised reinforcement learning method called surprise minimizing reinforcement learning (SMiRL). SMiRL alternates between learning a density model to evaluate the surprise of a stimulus, and improving the policy to seek more predictable stimuli. The policy seeks out stable and repeatable situations that counteract the environment's prevailing sources of entropy. This might include avoiding other hostile agents, or finding a stable, balanced pose for a bipedal robot in the face of disturbance forces. We demonstrate that our surprise minimizing agents can successfully play Tetris, Doom, control a humanoid to avoid falls, and navigate to escape enemies in a maze without any task-specific reward supervision. We further show that SMiRL can be used together with standard task rewards to accelerate reward-driven learning.",Reinforcement learning,oral presentations,1
28 Sept 2020,Evolving Reinforcement Learning Algorithms ,"We propose a method for meta-learning reinforcement learning algorithms by searching over the space of computational graphs which compute the loss function for a value-based model-free RL agent to optimize. The learned algorithms are domain-agnostic and can generalize to new environments not seen during training. Our method can both learn from scratch and bootstrap off known existing algorithms, like DQN, enabling interpretable modifications which improve performance. Learning from scratch on simple classical control and gridworld tasks, our method rediscovers the temporal-difference (TD) algorithm. Bootstrapped from DQN, we highlight two learned algorithms which obtain good generalization performance over other classical control tasks, gridworld type tasks, and Atari games. The analysis of the learned algorithm behavior shows resemblance to recently proposed RL algorithms that address overestimation in value-based methods.","reinforcement learning, evolutionary algorithms, meta-learning, genetic programming",oral presentations,1
28 Sept 2020,Growing Efficient Deep Networks by Structured Continuous Sparsification,"We develop an approach to growing deep network architectures over the course of training, driven by a principled combination of accuracy and sparsity objectives.  Unlike existing pruning or architecture search techniques that operate on full-sized models or supernet architectures, our method can start from a small, simple seed architecture and dynamically grow and prune both layers and filters.  By combining a continuous relaxation of discrete network structure optimization with a scheme for sampling sparse subnetworks, we produce compact, pruned networks, while also drastically reducing the computational expense of training.  For example, we achieve  inference FLOPs and  training FLOPs savings compared to a baseline ResNet-50 on ImageNet, while maintaining  top-1 validation accuracy --- all without any dedicated fine-tuning stage.  Experiments across CIFAR, ImageNet, PASCAL VOC, and Penn Treebank, with convolutional networks for image classification and semantic segmentation, and recurrent networks for language modeling, demonstrate that we both train faster and produce more efficient networks than competing architecture pruning or search methods.","deep learning, computer vision, network pruning, neural architecture search",oral presentations,1
28 Sept 2020,Deformable DETR: Deformable Transformers for End-to-End Object Detection,"DETR has been recently proposed to eliminate the need for many hand-designed components in object detection while demonstrating good performance. However, it suffers from slow convergence and limited feature spatial resolution, due to the limitation of Transformer attention modules in processing image feature maps. To mitigate these issues, we proposed Deformable DETR, whose attention modules only attend to a small set of key sampling points around a reference. Deformable DETR can achieve better performance than DETR (especially on small objects) with 10 less training epochs. Extensive experiments on the COCO benchmark demonstrate the effectiveness of our approach. Code is released at https://github.com/fundamentalvision/Deformable-DETR.","Efficient Attention Mechanism, Deformation Modeling, Multi-scale Representation, End-to-End Object Detection",oral presentations,1
28 Sept 2020,EigenGame: PCA as a Nash Equilibrium ,We present a novel view on principal components analysis as a competitive game in which each approximate eigenvector is controlled by a player whose goal is to maximize their own utility function. We analyze the properties of this PCA game and the behavior of its gradient based updates. The resulting algorithm---which combines elements from Oja's rule with a  generalized Gram-Schmidt orthogonalization---is naturally decentralized and hence parallelizable through message passing. We demonstrate the scalability of the algorithm with experiments on large image datasets and neural network activations. We discuss how this new view of PCA as a differentiable game can lead to further algorithmic developments and insights.,"pca, principal components analysis, nash, games, eigendecomposition, svd, singular value decomposition",oral presentations,1
28 Sept 2020,Augmenting Physical Models with Deep Networks for Complex Dynamics Forecasting,"Forecasting complex dynamical phenomena in settings where only partial knowledge of their dynamics is available is a prevalent problem across various scientific fields. While purely data-driven approaches are arguably insufficient in this context, standard physical modeling based approaches tend to be over-simplistic, inducing non-negligible errors. In this work, we introduce the APHYNITY framework, a principled approach for augmenting incomplete physical dynamics described by differential equations with deep data-driven models. It consists in decomposing the dynamics into two components: a physical component accounting for the dynamics for which we have some prior knowledge, and a data-driven component accounting for errors of the physical model. The learning problem is carefully formulated such that the physical model explains as much of the data as possible, while the data-driven component only describes information that cannot be captured by the physical model, no more, no less. This not only provides the existence and uniqueness for this decomposition, but also ensures interpretability and benefits generalization. Experiments made on three important use cases, each representative of a different family of phenomena, i.e. reaction-diffusion equations, wave equations and the non-linear damped pendulum, show that APHYNITY can efficiently leverage approximate physical models to accurately forecast the evolution of the system and correctly identify relevant physical parameters.","spatio-temporal forecasting, deep learning, physics, differential equations, hybrid systems",oral presentations,1
28 Sept 2020,Complex Query Answering with Neural Link Predictors,"Neural link predictors are immensely useful for identifying missing edges in large scale Knowledge Graphs. However, it is still not clear how to use these models for answering more complex queries that arise in a number of domains, such as queries using logical conjunctions (), disjunctions () and existential quantifiers (), while accounting for missing edges. In this work, we propose a framework for efficiently answering complex queries on incomplete Knowledge Graphs. We translate each query into an end-to-end differentiable objective, where the truth value of each atom is computed by a pre-trained neural link predictor. We then analyse two solutions to the optimisation problem, including gradient-based and combinatorial search. In our experiments, the proposed approach produces more accurate results than state-of-the-art methods --- black-box neural models trained on millions of generated queries --- without the need of training on a large and diverse set of complex queries. Using orders of magnitude less training data, we obtain relative improvements ranging from 8% up to 40% in Hits@3 across different knowledge graphs containing factual information. Finally, we demonstrate that it is possible to explain the outcome of our model in terms of the intermediate solutions identified for each of the complex query atoms. All our source code and datasets are available online, at https://github.com/uclnlp/cqd.","neural link prediction, complex query answering",oral presentations,1
28 Sept 2020,Towards Nonlinear Disentanglement in Natural Data with Temporal Sparse Coding,"Disentangling the underlying generative factors from complex data has so far been limited to carefully constructed scenarios. We propose a path towards natural data by first showing that the statistics of natural data provide enough structure to enable disentanglement, both theoretically and empirically. Specifically, we provide evidence that objects in natural movies undergo transitions that are typically small in magnitude with occasional large jumps, which is characteristic of a temporally sparse distribution. To address this finding we provide a novel proof that relies on a sparse prior on temporally adjacent observations to recover the true latent variables up to permutations and sign flips, directly providing a stronger result than previous work. We show that equipping practical estimation methods with our prior often surpasses the current state-of-the-art on several established benchmark datasets without any impractical assumptions, such as knowledge of the number of changing generative factors. Furthermore, we contribute two new benchmarks, Natural Sprites and KITTI Masks, which integrate the measured natural dynamics to enable disentanglement evaluation with more realistic datasets. We leverage these benchmarks to test our theory, demonstrating improved performance. We also identify non-obvious challenges for current methods in scaling to more natural domains. Taken together our work addresses key issues in disentanglement research for moving towards more natural settings. ","disentanglement, independent component analysis, natural scene statistics",oral presentations,1
28 Sept 2020,Self-training For Few-shot Transfer Across Extreme Task Differences,"Most few-shot learning techniques are pre-trained on a large, labeled “base dataset”. In problem domains where such large labeled datasets are not available for pre-training (e.g., X-ray, satellite images), one must resort to pre-training in a different “source” problem domain (e.g., ImageNet), which can be very different from the desired target task. Traditional few-shot and transfer learning techniques fail in the presence of such extreme differences between the source and target tasks. In this paper, we present a simple and effective solution to tackle this extreme domain gap: self-training a source domain representation on unlabeled data from the target domain. We show that this improves one-shot performance on the target domain by 2.9 points on average on the challenging BSCD-FSL benchmark consisting of datasets from multiple domains.","few-shot learning, self-training, cross-domain few-shot learning",oral presentations,1
28 Sept 2020,Score-Based Generative Modeling through Stochastic Differential Equations,"Creating noise from data is easy; creating data from noise is generative modeling. We present a stochastic differential equation (SDE) that smoothly transforms a complex data distribution to a known prior distribution by slowly injecting noise, and a corresponding reverse-time SDE that transforms the prior distribution back into the data distribution by slowly removing the noise. Crucially, the reverse-time SDE depends only on the time-dependent gradient field (a.k.a., score) of the perturbed data distribution. By leveraging advances in score-based generative modeling, we can accurately estimate these scores with neural networks, and use numerical SDE solvers to generate samples. We show that this framework encapsulates previous approaches in score-based generative modeling and diffusion probabilistic modeling, allowing for new sampling procedures and new modeling capabilities. In particular, we introduce a predictor-corrector framework to correct errors in the evolution of the discretized reverse-time SDE. We also derive an equivalent neural ODE that samples from the same distribution as the SDE, but additionally enables exact likelihood computation, and improved sampling efficiency. In addition, we provide a new way to solve inverse problems with score-based models, as demonstrated with experiments on class-conditional generation, image inpainting, and colorization. Combined with multiple architectural improvements, we achieve record-breaking performance for unconditional image generation on CIFAR-10 with an Inception score of 9.89 and FID of 2.20, a competitive likelihood of 2.99 bits/dim, and demonstrate high fidelity generation of  images for the first time from a score-based generative model.","generative models, score-based generative models, stochastic differential equations, score matching, diffusion",oral presentations,1
28 Sept 2020,Share or Not? Learning to Schedule Language-Specific Capacity for Multilingual Translation ,"Using a mix of shared and language-specific (LS) parameters has shown promise in multilingual neural machine  translation (MNMT), but the question of when and where LS capacity matters most is still under-studied. We offer such a study by proposing conditional language-specific routing (CLSR). CLSR employs hard binary gates conditioned on token representations to dynamically select LS or shared paths. By manipulating these gates, it can schedule LS capacity across sub-layers in MNMT subject to the guidance of translation signals and budget constraints. Moreover, CLSR can easily scale up to massively multilingual settings. Experiments with Transformer on OPUS-100 and WMT datasets show that: 1) MNMT is sensitive to both the amount and the position of LS modeling: distributing 10%-30% LS computation to the top and/or bottom encoder/decoder layers delivers the best performance; and 2) one-to-many translation benefits more from CLSR compared to many-to-one translation, particularly with unbalanced training data. Our study further verifies the trade-off between the shared capacity and LS capacity for multilingual translation. We corroborate our analysis by confirming the soundness of our findings as foundation of our improved multilingual Transformers. Source code and models are available at https://github.com/bzhangGo/zero/tree/iclr2021_clsr.","language-specific modeling, conditional computation, multilingual translation, multilingual transformer",oral presentations,1
28 Sept 2020,Image GANs meet Differentiable Rendering for Inverse Graphics and Interpretable 3D Neural Rendering ,"Differentiable rendering has paved the way to training neural networks to perform “inverse graphics” tasks such as predicting 3D geometry from monocular photographs. To train high performing models, most of the current approaches rely on multi-view imagery which are not readily available in practice.  Recent Generative Adversarial Networks (GANs) that synthesize images, in contrast, seem to acquire 3D knowledge implicitly during training: object viewpoints can be manipulated by simply manipulating the latent codes. However, these latent codes often lack further physical interpretation and thus GANs cannot easily be inverted to perform explicit 3D reasoning. In this paper, we aim to extract and disentangle 3D knowledge learned by generative models by utilizing differentiable renderers. Key to our approach is to exploit GANs as a multi-view data generator to train an inverse graphics network using an off-the-shelf differentiable renderer, and the trained inverse graphics network as a teacher to disentangle the GAN's latent code into interpretable 3D properties. The entire architecture is trained iteratively using cycle consistency losses. We show that our approach significantly outperforms state-of-the-art inverse graphics networks trained on existing datasets, both quantitatively and via user studies. We further showcase the disentangled GAN as a controllable 3D “neural renderer"", complementing traditional graphics renderers.","Differentiable rendering, inverse graphics, GANs",oral presentations,1
28 Sept 2020,How Neural Networks Extrapolate: From Feedforward to Graph Neural Networks,"We study how neural networks trained by gradient descent  extrapolate, i.e., what they learn outside the support of the training distribution. Previous works report mixed empirical results when extrapolating with neural networks: while feedforward neural networks, a.k.a. multilayer perceptrons (MLPs), do not extrapolate well in certain simple tasks, Graph Neural Networks (GNNs) -- structured networks with MLP modules -- have shown some success in more complex tasks.  Working towards a theoretical explanation, we identify conditions under which MLPs and GNNs extrapolate well. First, we quantify the observation that ReLU MLPs quickly converge to linear functions along any direction from the origin, which implies that ReLU MLPs do not extrapolate most nonlinear functions. But, they can provably learn a linear target function when the training distribution is sufficiently diverse. Second, in connection to analyzing the successes and limitations of GNNs, these results suggest a hypothesis for which we provide theoretical and empirical evidence: the success of GNNs in extrapolating algorithmic tasks to new data (e.g., larger graphs or edge weights) relies on encoding task-specific non-linearities in the architecture or features. Our theoretical analysis builds on a connection of over-parameterized networks to the neural tangent kernel.  Empirically, our theory holds across different training settings.","extrapolation, deep learning, out-of-distribution, graph neural networks, deep learning theory",oral presentations,1
28 Sept 2020,Contrastive Explanations for Reinforcement Learning via Embedded Self Predictions,"We investigate a deep reinforcement learning (RL) architecture that supports explaining why a learned agent prefers one action over another. The key idea is to learn action-values that are directly represented via human-understandable properties of expected futures. This is realized via the embedded self-prediction (ESP) model, which learns said properties in terms of human provided features. Action preferences can then be explained by contrasting the future properties predicted for each action. To address cases where there are a large number of features, we develop a novel method for computing minimal sufficient explanations from an ESP. Our case studies in three domains, including a complex strategy game, show that ESP models can be effectively learned and support insightful explanations. ","Explainable AI, Deep Reinforcement Learning",oral presentations,1
28 Sept 2020,Improved Autoregressive Modeling with Distribution Smoothing,"While autoregressive models excel at image compression, their sample quality is often lacking. Although not realistic, generated images often have high likelihood according to the model, resembling the case of adversarial examples. Inspired by a successful adversarial defense method, we incorporate randomized smoothing into autoregressive generative modeling. We first model a smoothed version of the data distribution, and then reverse the smoothing process to recover the original data distribution. This procedure drastically improves the sample quality of existing autoregressive models on several synthetic and real-world image datasets while obtaining competitive likelihoods on synthetic datasets.","generative models, autoregressive models",oral presentations,1
28 Sept 2020,MONGOOSE: A Learnable LSH Framework for Efficient Neural Network Training ,"Recent advances by practitioners in the deep learning community have breathed new life into Locality Sensitive Hashing (LSH), using it to reduce memory and time bottlenecks in neural network (NN) training. However, while LSH has sub-linear guarantees for approximate near-neighbor search in theory, it is known to have inefficient query time in practice due to its use of random hash functions. Moreover, when model parameters are changing, LSH suffers from update overhead. This work is motivated by an observation that model parameters evolve slowly, such that the changes do not always require an LSH update to maintain performance. This phenomenon points to the potential for a reduction in update time and allows for a modified learnable version of data-dependent LSH to improve query time at a low cost. We use the above insights to build MONGOOSE, an end-to-end LSH framework for efficient NN training. In particular, MONGOOSE is equipped with a scheduling algorithm to adaptively perform LSH updates with provable guarantees and learnable hash functions to improve query efficiency. Empirically, we validate MONGOOSE on large-scale deep learning models for recommendation systems and language modeling. We find that it achieves up to 8% better accuracy compared to previous LSH approaches, with  speed-up and  reduction in memory usage.","Large-scale Deep Learning, Large-scale Machine Learning, Efficient Training, Randomized Algorithms",oral presentations,1
28 Sept 2020,Gradient Projection Memory for Continual Learning,"The ability to learn continually without forgetting the past tasks is a desired attribute for artificial learning systems. Existing approaches to enable such learning in artificial neural networks usually rely on network growth, importance based weight update or replay of old data from the memory. In contrast, we propose a novel approach where a neural network learns new tasks by taking gradient steps in the orthogonal direction to the gradient subspaces deemed important for the past tasks. We find the bases of these subspaces by analyzing network representations (activations) after learning each task with Singular Value Decomposition (SVD) in a single shot manner and store them in the memory as Gradient Projection Memory (GPM). With qualitative and quantitative analyses, we show that such orthogonal gradient descent induces minimum to no interference with the past tasks, thereby mitigates forgetting. We evaluate our algorithm on diverse image classification datasets with short and long sequences of tasks and report better or on-par performance compared to the state-of-the-art approaches. ","Continual Learning, Representation Learning, Computer Vision, Deep learning",oral presentations,1
28 Sept 2020,Why Are Convolutional Nets More Sample-Efficient than Fully-Connected Nets?,"Convolutional neural networks often dominate fully-connected counterparts in generalization performance, especially on image classification tasks. This is often explained in terms of \textquotedblleft better inductive bias.\textquotedblright\  However, this has not been made mathematically rigorous, and the hurdle is that the sufficiently wide fully-connected net can always simulate the convolutional net. Thus the training algorithm plays a role. The current work describes a natural task on which a provable sample complexity gap can be shown, for standard training algorithms. We construct a single natural distribution on  on which any orthogonal-invariant algorithm (i.e. fully-connected networks trained with most gradient-based methods from gaussian initialization) requires  samples to generalize while  samples suffice for convolutional architectures. Furthermore, we demonstrate a single target function, learning which on all possible distributions leads to an  vs  gap. The proof relies on the fact that SGD on fully-connected network is orthogonal equivariant. Similar results are achieved for  regression and adaptive training algorithms, e.g. Adam and AdaGrad, which are only permutation equivariant.","sample complexity separation, equivariance, convolutional neural networks, fully-connected",oral presentations,1
28 Sept 2020,Iterated learning for emergent systematicity in VQA ,"Although neural module networks have an architectural bias towards compositionality, they require gold standard layouts to generalize systematically in practice. When instead learning layouts and modules jointly, compositionality does not arise automatically and an explicit pressure is necessary for the emergence of layouts exhibiting the right structure. We propose to address this problem using iterated learning, a cognitive science theory of the emergence of compositional languages in nature that has primarily been applied to simple referential games in machine learning. Considering the layouts of module networks as samples from an emergent language, we use iterated learning to encourage the development of structure within this language. We show that the resulting layouts support systematic generalization in neural agents solving the more complex task of visual question-answering. Our regularized iterated learning method can outperform baselines without iterated learning on SHAPES-SyGeT (SHAPES Systematic Generalization Test), a new split of the SHAPES dataset we introduce to evaluate systematic generalization, and on CLOSURE, an extension of CLEVR also designed to test systematic generalization. We demonstrate superior performance in recovering ground-truth compositional program structure with limited supervision on both SHAPES-SyGeT and CLEVR.","iterated learning, cultural transmission, neural module network, clevr, shapes, vqa, visual question answering, systematic generalization, compositionality",oral presentations,1
28 Sept 2020,Coupled Oscillatory Recurrent Neural Network (coRNN): An accurate and (gradient) stable architecture for learning long time dependencies ,"Circuits of biological neurons, such as in the functional parts of the brain can be modeled as networks of coupled oscillators. Inspired by the ability of these systems to express a rich set of outputs while keeping (gradients of) state variables bounded, we propose a novel architecture for recurrent neural networks. Our proposed RNN is based on a time-discretization of a system of second-order ordinary differential equations, modeling networks of controlled nonlinear oscillators. We prove precise bounds on the gradients of the hidden states, leading to the mitigation of the exploding and vanishing gradient problem for this RNN. Experiments show that the proposed RNN is comparable in performance to the state of the art on a variety of benchmarks, demonstrating the potential of this architecture to provide stable and accurate RNNs for processing complex sequential data.","RNNs, Oscillators, Gradient stability, Long-term dependencies",oral presentations,1
28 Sept 2020,Toward Trainability of Quantum Neural Networks ,"Quantum Neural Networks (QNNs) have been recently proposed as generalizations of classical neural networks to achieve the quantum speed-up. Despite the potential to outperform classical models, serious bottlenecks exist for training QNNs; namely, QNNs with random structures have poor trainability due to the vanishing gradient with rate exponential to the input qubit number. The vanishing gradient could seriously influence the applications of large-size QNNs. In this work, we provide the first viable solution with theoretical guarantees. Specifically, we prove that QNNs with tree tensor and step controlled architectures have gradients that vanish at most polynomially with the qubit number. Moreover, our result holds irrespective of which encoding methods are employed. We numerically demonstrate QNNs with tree tensor and step controlled structures for the application of binary classification. Simulations show faster convergent rates and better accuracy compared to QNNs with random structures.","Near-term Quantum Algorithm, Quantum Neural Network, Trainability, Hierarchical Structure",rejected,0
28 Sept 2020,Selecting Treatment Effects Models for Domain Adaptation Using Causal Knowledge,"Selecting causal inference models for estimating individualized treatment effects (ITE) from observational data presents a unique challenge since the counterfactual outcomes are never observed. The problem is challenged further in the unsupervised domain adaptation (UDA) setting where we only have access to labeled samples in the source domain, but desire selecting a model that achieves good performance on a target domain for which only unlabeled samples are available. Existing techniques for UDA model selection are designed for the predictive setting. These methods examine discriminative density ratios between the input covariates in the source and target domain and do not factor in the model's predictions in the target domain. Because of this, two models with identical performance on the source domain would receive the same risk score by existing methods, but in reality, have significantly different performance in the test domain. We leverage the invariance of causal structures across domains to propose a novel model selection metric specifically designed for ITE methods under the UDA setting. In particular, we propose selecting models whose predictions of interventions' effects satisfy known causal structures in the target domain. Experimentally, our method selects ITE models that are more robust to covariate shifts on several healthcare datasets, including estimating the effect of ventilation in COVID-19 patients from different geographic locations.","causal inference, treatment effects, healthcare",rejected,0
28 Sept 2020,Closing the Generalization Gap in One-Shot Object Detection,"Despite substantial progress in object detection and few-shot learning, detecting objects based on a single example - one-shot object detection - remains a challenge. A central problem is the generalization gap: Object categories used during training are detected much more reliably than novel ones. We here show that this generalization gap can be nearly closed by increasing the number of object categories used during training. Doing so allows us to beat the state-of-the-art on COCO by 5.4 %AP50 (from 22.0 to 27.5) and improve generalization from seen to unseen classes from 45% to 89%. We verify that the effect is caused by the number of categories and not the amount of data and that it holds for different models, backbones and datasets. This result suggests that the key to strong few-shot detection models may not lie in sophisticated metric learning approaches, but instead simply in scaling the number of categories. We hope that our findings will help to better understand the challenges of few-shot learning and encourage future data annotation efforts to focus on wider datasets with a broader set of categories rather than gathering more samples per category.","One-Shot Learning, Few-Shot Learning, Object Detection, One-Shot Object Detection, Generalization",rejected,0
28 Sept 2020,PettingZoo: Gym for Multi-Agent Reinforcement Learning ,"OpenAI's Gym library  contains a large, diverse set of environments that are useful benchmarks in reinforcement learning, under a single elegant Python API (with tools to develop new compliant environments) . The introduction of this library has proven a watershed moment for the reinforcement learning community, because it created an accessible set of benchmark environments that everyone could use (including wrapper important existing libraries), and because a standardized API let RL learning methods and environments from anywhere be trivially exchanged. This paper similarly introduces PettingZoo, a library of diverse set of multi-agent environments under a single elegant Python API, with tools to easily make new compliant environments.","Reinforcement Learning, Multi-agent Reinforcement Learning",rejected,0
28 Sept 2020,Towards Practical Second Order Optimization for Deep Learning ,"Optimization in machine learning, both theoretical and applied, is presently dominated by first-order gradient methods such as stochastic gradient descent. Second-order optimization methods, that involve second derivatives and/or second order statistics of the data, are far less prevalent despite strong theoretical properties, due to their prohibitive computation, memory and communication costs.   In an attempt to bridge this gap between theoretical and practical optimization, we present a scalable implementation of a second-order preconditioned method (concretely, a variant of full-matrix Adagrad), that along with several critical algorithmic and numerical improvements, provides significant convergence and wall-clock time improvements compared to conventional first-order methods on state-of-the-art deep models. Our novel design effectively utilizes the prevalent heterogeneous hardware architecture for training deep models, consisting of a multicore CPU coupled with multiple accelerator units. We demonstrate superior performance compared to state-of-the-art on very large learning tasks such as machine translation with Transformers, language modeling with BERT, click-through rate prediction on Criteo, and image classification on ImageNet with ResNet-50.","large scale distributed deep learning, second order optimization, bert, resnet, criteo, transformer, machine translation",rejected,0
28 Sept 2020,Transforming Recurrent Neural Networks with Attention and Fixed-point Equations ,"Transformer has achieved state of the art performance in multiple Natural Language Processing tasks recently. Yet the Feed Forward Network(FFN) in a Transformer block is computationally expensive. In this paper, we present a framework to transform Recurrent Neural Networks(RNNs) and their variants into self-attention-style models, with an approximation of Banach Fixed-point Theorem. Within this framework, we propose a new model, StarSaber, by solving a set of equations obtained from RNN with Fixed-point Theorem and further approximate it with a Multi-layer Perceptron. It provides a view of stacking layers. StarSaber achieves better performance than both the vanilla Transformer and an improved version called ReZero on three datasets and is more computationally efficient, due to the reduction of Transformer's FFN layer. It has two major parts. One is a way to encode position information with two different matrices. For every position in a sequence, we have a matrix operating on positions before it and another matrix operating on positions after it. The other is the introduction of direct paths from the input layer to the rest of layers. Ablation studies show the effectiveness of these two parts. We additionally show that other RNN variants such as RNNs with gates can also be transformed in the same way, outperforming the two kinds of Transformers as well.","Fixed-point, Attention, Feed Forward Network, Transformer, Recurrent Neural Network, Deep Learning",rejected,0
28 Sept 2020,Truthful Self-Play ,"We present a general framework for evolutionary learning to emergent unbiased state representation without any supervision. Evolutionary frameworks such as self-play converge to bad local optima in case of multi-agent reinforcement learning in non-cooperative partially observable environments with communication due to information asymmetry.  Our proposed framework is a simple modification of self-play inspired by mechanism design, also known as {\em reverse game theory}, to elicit truthful signals and make the agents cooperative. The key idea is to add imaginary rewards using the peer prediction method, i.e., a mechanism for evaluating the validity of information exchanged between agents in a decentralized environment. Numerical experiments with predator prey, traffic junction and StarCraft tasks demonstrate that the state-of-the-art performance of our framework.","Comm-POSG, Imaginary Rewards",rejected,0
28 Sept 2020,Segmenting Natural Language Sentences via Lexical Unit Analysis,"In this work, we present Lexical Unit Analysis (LUA), a framework for general sequence segmentation tasks. Given a natural language sentence, LUA scores all the valid segmentation candidates and utilizes dynamic programming (DP) to extract the maximum scoring one. LUA enjoys a number of appealing properties such as inherently guaranteeing the predicted segmentation to be valid and facilitating globally optimal training and inference. Besides, the practical time complexity of LUA can be reduced to linear time, which is very efficient. We have conducted extensive experiments on 5 tasks, including syntactic chunking, named entity recognition (NER), slot filling, Chinese word segmentation, and Chinese part-of-speech (POS) tagging, across 15 datasets. Our models have achieved the state-of-the-art performances on 13 of them. The results also show that the F1 score of identifying long-length segments is notably improved.","Neural Sequence Labeling, Neural Sequence Segmentation, Dynamic Programming",rejected,0
28 Sept 2020,Mixture of Step Returns in Bootstrapped DQN ,"The concept of utilizing multi-step returns for updating value functions has been adopted in deep reinforcement learning (DRL) for a number of years. Updating value functions with different backup lengths provides advantages in different aspects, including bias and variance of value estimates, convergence speed, and exploration behavior of the agent. Conventional methods such as TD-lambda leverage these advantages by using a target value equivalent to an exponential average of different step returns. Nevertheless, integrating step returns into a single target sacrifices the diversity of the advantages offered by different step return targets. To address this issue, we propose Mixture Bootstrapped DQN (MB-DQN) built on top of bootstrapped DQN, and uses different backup lengths for different bootstrapped heads. MB-DQN enables heterogeneity of the target values that is unavailable in approaches relying only on a single target value. As a result, it is able to maintain the advantages offered by different backup lengths. In this paper, we first discuss the motivational insights through a simple maze environment. In order to validate the effectiveness of MB-DQN, we perform experiments on the Atari 2600 benchmark environments and demonstrate the performance improvement of MB-DQN over a number of baseline methods. We further provide a set of ablation studies to examine the impacts of different design configurations of MB-DQN.",Reinforcement Learning,rejected,0
28 Sept 2020,THE EFFICACY OF L1 REGULARIZATION IN NEURAL NETWORKS ,"A crucial problem in neural networks is to select the most appropriate number of hidden neurons and obtain tight statistical risk bounds. In this work, we present a new perspective towards the bias-variance tradeoff in neural networks. As an alternative to selecting the number of neurons, we theoretically show that  regularization can control the generalization error and sparsify the input dimension. In particular, with an appropriate  regularization on the output layer, the network can produce a statistical risk that is near minimax optimal. Moreover, an appropriate  regularization on the input layer leads to a risk bound that does not involve the input data dimension. Our analysis is based on a new amalgamation of dimension-based and norm-based complexity analysis to bound the generalization error. A consequent observation from our results is that an excessively large number of neurons do not necessarily inflate generalization errors under a suitable regularization.","Model selection, Neural Network, Regularization",rejected,0
28 Sept 2020,Proper Measure for Adversarial Robustness,"This paper analyzes the problems of adversarial accuracy and adversarial training. We argue that standard adversarial accuracy fails to properly measure the robustness of classifiers. Its definition has a tradeoff with standard accuracy even when we neglect generalization. In order to handle the problems of the standard adversarial accuracy, we introduce a new measure for the robustness of classifiers called genuine adversarial accuracy. It can measure the adversarial robustness of classifiers without trading off accuracy on clean data and accuracy on the adversarially perturbed samples. In addition, it does not favor a model with invariance-based adversarial examples, samples whose predicted classes are unchanged even if the perceptual classes are changed. We prove that a single nearest neighbor (1-NN) classifier is the most robust classifier according to genuine adversarial accuracy for given data and a norm-based distance metric when the class for each data point is unique. Based on this result, we suggest that using poor distance metrics might be one factor for the tradeoff between test accuracy and  norm-based test adversarial robustness.","adversarial examples, adversarial robustness, adversarial accuracy, nearest neighbor classifiers",rejected,0
28 Sept 2020,Meta-k: Towards Unsupervised Prediction of Number of Clusters,"Data clustering is a well-known unsupervised learning approach. Despite the recent advances in clustering using deep neural networks, determining the number of clusters without any information about the given dataset remains an existing problem. There have been classical approaches based on data statistics that require the manual analysis of a data scientist to calculate the probable number of clusters in a dataset. In this work, we propose a new method for unsupervised prediction of the number of clusters in a dataset given only the data without any labels. We evaluate our method extensively on randomly generated datasets using the scikit-learn package and multiple computer vision datasets and show that our method is able to determine the number of classes in a dataset effectively without any supervision.","Clustering, Self-supervised learning, Meta-learning",rejected,0
28 Sept 2020,Transformers satisfy,"The Propositional Satisfiability Problem (SAT), and more generally, the Constraint Satisfaction Problem (CSP), are mathematical questions defined as finding an assignment to a set of objects that satisfies a series of constraints. The modern approach is trending to solve CSP through neural symbolic methods. Most recent works are sequential model-based and adopt neural embedding, i.e., reinforcement learning with neural graph networks, and graph recurrent neural networks. This work proposes a one-shot model derived from the eminent Transformer architecture for factor graph structure to solve the CSP problem. We define the heterogeneous attention mechanism based on meta-paths for the self-attention between literals, the cross-attention based on the bipartite graph links from literal to clauses, or vice versa. This model takes advantage of parallelism. Our model achieves high speed and very high accuracy on the factor graph for CSPs with arbitrary size.","constraint satisfaction problem, graph attention, transformers",rejected,0
28 Sept 2020,Stochastic Inverse Reinforcement Learning,"The goal of the inverse reinforcement learning (IRL) problem is to recover the reward functions from expert demonstrations. However, the IRL problem like any ill-posed inverse problem suffers the congenital defect that the policy may be optimal for many reward functions, and expert demonstrations may be optimal for many policies. In this work, we generalize the IRL problem to a well-posed expectation optimization problem stochastic inverse reinforcement learning (SIRL) to recover the probability distribution over reward functions. We adopt the Monte Carlo expectation-maximization (MCEM) method to estimate the parameter of the probability distribution as the first solution to the SIRL problem. The solution is succinct, robust, and transferable for a learning task and can generate alternative solutions to the IRL problem. Through our formulation, it is possible to observe the intrinsic property for the IRL problem from a global viewpoint, and our approach achieves a considerable performance on the objectworld. ","Inverse Reinforcement Learning, Stochastic Methods, MCEM",rejected,0
28 Sept 2020,GN-Transformer: Fusing AST and Source Code information in Graph Networks,"As opposed to natural languages, source code understanding is influenced by grammar relations between tokens regardless of their identifier name. Considering graph representation of source code such as Abstract Syntax Tree (AST) and Control Flow Graph (CFG), can capture a token’s grammatical relationships that are not obvious from the source code. Most existing methods are late fusion and underperform when supplementing the source code text with a graph representation. We propose a novel method called GN-Transformer to fuse representations learned from graph and text modalities under the Graph Networks (GN) framework with attention mechanism. Our method learns the embedding on a constructed graph called Syntax-Code Graph (SCG). We perform experiments on the structure of SCG, an ablation study on the model design and the hyper-paramaters to conclude that the performance advantage is from the fusion method and not the specific details of the model. The proposed method achieved state of the art performance in two code summarization datasets and across three metrics.",,rejected,0
28 Sept 2020,Hierarchical Meta Reinforcement Learning for Multi-Task Environments,"Deep reinforcement learning algorithms aim to achieve human-level intelligence by solving practical decisions-making problems, which are often composed of multiple sub-tasks. Complex and subtle relationships between sub-tasks make traditional methods hard to give a promising solution. We implement a first-person shooting environment with random spatial structures to illustrate a typical representative of this kind. A desirable agent should be capable of balancing between different sub-tasks: navigation to find enemies and shooting to kill them. To address the problem brought by the environment, we propose a Meta Soft Hierarchical reinforcement learning framework (MeSH), in which each low-level sub-policy focuses on a specific sub-task respectively and high-level policy automatically learns to utilize low-level sub-policies through meta-gradients. The proposed framework is able to disentangle multiple sub-tasks and discover proper low-level policies under different situations. The effectiveness and efficiency of the framework are shown by a series of comparison experiments. Both environment and algorithm code will be provided for open source to encourage further research.","Reinforcement Learning, Multi-task, Hierarchical, Meta Learning",rejected,0
28 Sept 2020,Continuous Transfer Learning ,"Transfer learning has been successfully applied across many high-impact applications. However, most existing work focuses on the static transfer learning setting, and very little is devoted to modeling the time evolving target domain, such as the online reviews for movies. To bridge this gap, in this paper, we focus on the continuous transfer learning setting with a time evolving target domain. One major challenge associated with continuous transfer learning is the time evolving relatedness of the source domain and the current target domain as the target domain evolves over time. To address this challenge, we first derive a generic generalization error bound on the current target domain with flexible domain discrepancy measures. Furthermore, a novel label-informed C-divergence is proposed to measure the shift of joint data distributions (over input features and output labels) across domains. It could be utilized to instantiate a tighter error upper bound in the continuous transfer learning setting, thus motivating us to develop an adversarial Variational Auto-encoder algorithm named CONTE by minimizing the C-divergence based error upper bound. Extensive experiments on various data sets demonstrate the effectiveness of our CONTE algorithm.",,rejected,0
28 Sept 2020,Patch-level Neighborhood Interpolation: A General and Effective Graph-based Regularization Strategy,"Regularization plays a crucial role in machine learning models, especially for deep neural networks. The existing regularization techniques mainly rely on the i.i.d. assumption and only consider the knowledge from the current sample, without the leverage of the neighboring relationship between samples. In this work, we propose a general regularizer called Patch-level Neighborhood Interpolation~(\textbf{Pani}) that conducts a non-local representation in the computation of network. Our proposal explicitly constructs patch-level graphs in different network layers and then linearly interpolates neighborhood patch features, serving as a general and effective regularization strategy. Further, we customize our approach into two kinds of popular regularization methods, namely Virtual Adversarial Training (VAT) and MixUp as well as its variants. The first derived \textbf{Pani VAT} presents a novel way to construct non-local adversarial smoothness by employing patch-level interpolated perturbations. In addition, the second derived \textbf{Pani MixUp} method extends the original MixUp regularization and its variant to the Pani version, achieving a  significant improvement in the performance. Finally, extensive experiments are conducted to verify the effectiveness of our Patch-level Neighborhood Interpolation approach in both supervised and semi-supervised settings.","Deep Learning, Regularization, Graph-based Representation",rejected,0
28 Sept 2020,Grounded Compositional Generalization with Environment Interactions,"In this paper, we present a compositional generalization approach in grounded agent instruction learning. Compositional generalization is an important part of human intelligence, but current neural network models do not have such ability. This is more complicated in multi-modal problems with grounding. Our proposed approach has two main ideas. First, we use interactions between agent and the environment to find components in the output. Second, we apply entropy regularization to learn corresponding input components for each output component. The results show the proposed approach significantly outperforms baselines in most tasks, with more than 25% absolute average accuracy increase. We also investigate the impact of entropy regularization and other changes with ablation study. We hope this work is the first step to address grounded compositional generalization, and it will be helpful in advancing artificial intelligence research.","compositional generalization, grounding",rejected,0
28 Sept 2020,Secure Network Release with Link Privacy ,"Many data mining and analytical tasks rely on the abstraction of networks (graphs) to summarize relational structures among individuals (nodes). Since relational data are often sensitive, we aim to seek effective approaches to release utility-preserved yet privacy-protected structured data. In this paper, we leverage the differential privacy (DP) framework, to formulate and enforce rigorous privacy constraints on deep graph generation models, with a focus on edge-DP to guarantee individual link privacy. In particular, we enforce edge-DP by injecting Gaussian noise to the gradients of a link reconstruction based graph generation model, and ensure data utility by improving structure learning with structure-oriented graph comparison. Extensive experiments on two real-world network datasets show that our proposed DPGGAN model is able to generate networks with effectively preserved global structure and rigorously protected individual link privacy.","generative model, graph neural network, data release",rejected,0
28 Sept 2020,FLAG: Adversarial Data Augmentation for Graph Neural Networks ,"Data augmentation helps neural networks generalize better, but it remains an open question how to effectively augment graph data to enhance the performance of GNNs (Graph Neural Networks). While most existing graph regularizers focus on augmenting graph topological structures by adding/removing edges, we offer a novel direction to augment in the input node feature space for better performance. We propose a simple but effective solution, FLAG (Free Large-scale Adversarial Augmentation on Graphs), which iteratively augments node features with gradient-based adversarial perturbations during training, and boosts performance at test time. Empirically, FLAG can be easily implemented with a dozen lines of code and is flexible enough to function with any GNN backbone, on a wide variety of large-scale datasets, and in both transductive and inductive settings. Without modifying a model's architecture or training setup, FLAG yields a consistent and salient performance boost across both node and graph classification tasks. Using FLAG, we reach state-of-the-art performance on the large-scale ogbg-molpcba, ogbg-ppa, and ogbg-code datasets.  ","Graph Neural Networks, Data Augmentation, Adversarial Training",rejected,0
28 Sept 2020,Non-Inherent Feature Compatible Learning,"The need of Feature Compatible Learning (FCL) arises from many large scale retrieval-based applications, where updating the entire library of embedding vectors is expensive. When an upgraded embedding model shows potential, it is desired to transform the benefit of the new model without refreshing the library. While progresses have been made along this new direction, existing approaches for feature compatible learning mostly rely on old training data and classifiers, which are not available in many industry settings. In this work, we introduce an approach for feature compatible learning without inheriting old classifier and training data, i.e., Non-Inherent Feature Compatible Learning. Our approach requires only features extracted by \emph{old} model's backbone and \emph{new} training data, and makes no assumption about the overlap between old and new training data. We propose a unified framework for FCL, and extend it to handle the case where the old model is a black-box.  Specifically, we learn a simple pseudo classifier in lieu of the old model, and further enhance it with a random walk algorithm.  As a result, the embedding features produced by the new model can be matched with those from the old model without sacrificing performance. Experiments on ImageNet ILSVRC 2012 and Places365 data proved the efficacy of the proposed approach.","Deep Learning, Feature Learning, Compatible Learning",rejected,0
28 Sept 2020,Graph View-Consistent Learning Network,"Recent years, methods based on neural networks have made great achievements in solving large and complex graph problems. However, high efficiency of these methods depends on large training and validation sets, while the acquisition of ground-truth labels is expensive and time-consuming. In this paper, a graph view-consistent learning network (GVCLN) is specially designed for the semi-supervised learning when the number of the labeled samples is very small. We fully exploit the neighborhood aggregation capability of GVCLN and use dual views to obtain different representations. Although the two views have different viewing angles, their observation objects are the same, so their observation representations need to be consistent. For view-consistent representations between two views, two loss functions are designed besides a supervised loss. The supervised loss uses the known labeled set, while a view-consistent loss is applied to the two views to obtain the consistent representation and a pseudo-label loss is designed by using the common high-confidence predictions. GVCLN with these loss functions can obtain the view-consistent representations of the original feature. We also find that preprocessing the node features with specific filter before training is good for subsequent classification tasks. Related experiments have been done on the three citation network datasets of Cora, Citeseer, and PubMed. On several node classification tasks, GVCLN achieves state-of-the-art performance.",,rejected,0
28 Sept 2020,Self-Pretraining for Small Datasets by Exploiting Patch Information," Deep learning tasks with small datasets are often tackled by pretraining models with large datasets on relevent tasks. Although pretraining methods mitigate the problem of overfitting, it can be difficult to find appropriate pretrained models sometimes. In this paper, we proposed a self-pretraininng method by exploiting patch information in the dataset itself without pretraining on other datasets. Our experiments show that the self-pretraining method leads to better performance than training from scratch both in the condition of not using other data.","Learning with Small Datasets, Self-Pretraining",rejected,0
28 Sept 2020,On the Marginal Regret Bound Minimization of Adaptive Methods ,"Numerous adaptive algorithms such as AMSGrad and Radam have been proposed and applied to deep learning recently. However, these modifications do not improve the convergence rate of adaptive algorithms and whether a better algorithm exists still remains an open question. In this work, we propose a new motivation for designing the proximal function of adaptive algorithms, named as marginal regret bound minimization. Based on such an idea, we propose a new class of adaptive algorithms that not only achieves marginal optimality but can also potentially converge much faster than any existing adaptive algorithms in the long term. We show the superiority of the new class of adaptive algorithms both theoretically and empirically using experiments in deep learning. ","Optimization Algorithm, Adaptive algorithms, Online Learning, Regret Minimization",rejected,0
28 Sept 2020,SVMax: A Feature Embedding Regularizer,"A neural network regularizer (eg, weight decay) boosts performance by explicitly penalizing the complexity of a network. In this paper, we penalize inferior network activations -- feature embeddings -- which in turn regularize the network's weights implicitly. We propose singular value maximization (SVMax) to learn a uniform feature embedding. The SVMax regularizer integrates seamlessly with both supervised and unsupervised learning. During training, our formulation mitigates model collapse and enables larger learning rates. Thus, our formulation converges in fewer epochs, which reduces the training computational cost. We evaluate the SVMax regularizer using both retrieval and generative adversarial networks. We leverage a synthetic mixture of Gaussians dataset to evaluate SVMax in an unsupervised setting. For retrieval networks, SVMax achieves significant improvement margins across various ranking losses.","metric learning, model collapse, feature embedding, neural network regularizer",rejected,0
28 Sept 2020,Transferability of Compositionality,"Compositional generalization is the algebraic capacity to understand and produce large amount of novel combinations from known components. It is a key element of human intelligence for out-of-distribution generalization. To equip neural networks with such ability, many algorithms have been proposed to extract compositional representations from the training distribution. However, it has not been discussed whether the trained model can still extract such representations in the test distribution. In this paper, we argue that the extraction ability does not transfer naturally, because the extraction network suffers from the divergence of distributions. To address this problem, we propose to use an auxiliary reconstruction network with regularized hidden representations as input, and optimize the representations during inference. The proposed approach significantly improves accuracy, showing more than a 20% absolute increase in various experiments compared with baselines. To our best knowledge, this is the first work to focus on the transferability of compositionality, and it is orthogonal to existing efforts of learning compositional representations in training distribution. We hope this work will help to advance compositional generalization and artificial intelligence research.",Compositionality,rejected,0