sentence
stringlengths 373
5.09k
| label
stringclasses 2
values |
|---|---|
Title: A Generalized Training Approach for Multiagent Learning. Abstract: This paper investigates a population-based training regime based on game-theoretic principles called Policy-Spaced Response Oracles (PSRO). PSRO is general in the sense that it (1) encompasses well-known algorithms such as fictitious play and double oracle as special cases, and (2) in principle applies to general-sum, many-player games. Despite this, prior studies of PSRO have been focused on two-player zero-sum games, a regime where in Nash equilibria are tractably computable. In moving from two-player zero-sum games to more general settings, computation of Nash equilibria quickly becomes infeasible. Here, we extend the theoretical underpinnings of PSRO by considering an alternative solution concept, α-Rank, which is unique (thus faces no equilibrium selection issues, unlike Nash) and applies readily to general-sum, many-player settings. We establish convergence guarantees in several games classes, and identify links between Nash equilibria and α-Rank. We demonstrate the competitive performance of α-Rank-based PSRO against an exact Nash solver-based PSRO in 2-player Kuhn and Leduc Poker. We then go beyond the reach of prior PSRO applications by considering 3- to 5-player poker games, yielding instances where α-Rank achieves faster convergence than approximate Nash solvers, thus establishing it as a favorable general games solver. We also carry out an initial empirical validation in MuJoCo soccer, illustrating the feasibility of the proposed approach in another complex domain. | accept |
Title: Few-Round Learning for Federated Learning. Abstract: Federated learning (FL) presents an appealing opportunity for individuals who are willing to make their private data available for building a communal model without revealing their data contents to anyone else. Of central issues that may limit a widespread adoption of FL is the significant communication resources required in the exchange of updated model parameters between the server and individual clients over many communication rounds. In this work, we focus on limiting the number of model exchange rounds in FL to some small fixed number $R$, to control the communication burden. Following the spirit of meta-learning for few-shot learning, we take a meta-learning strategy to train the model so that once the meta-training phase is over, only $R$ rounds of FL would produce a model that will satisfy the needs of all participating clients. A key advantage of employing meta-training is that the main labeled dataset used in training could differ significantly (e.g., different classes of images) from the actual data sample presented at inference time. Compared to the meta-training approaches to optimize personalized local models at distributed devices, our method better handles the potential lack of data variability at individual nodes. Extensive experimental results indicate that meta-training geared to few-round learning provide large performance improvements compared to various baselines. | reject |
Title: Training independent subnetworks for robust prediction. Abstract: Recent approaches to efficiently ensemble neural networks have shown that strong robustness and uncertainty performance can be achieved with a negligible gain in parameters over the original network. However, these methods still require multiple forward passes for prediction, leading to a significant runtime cost. In this work, we show a surprising result:
the benefits of using multiple predictions can be achieved 'for free' under a single model's forward pass. In particular, we show that, using a multi-input multi-output (MIMO) configuration, one can utilize a single model's capacity to train multiple subnetworks that independently learn the task at hand. By ensembling the predictions made by the subnetworks, we improve model robustness without increasing compute. We observe a significant improvement in negative log-likelihood, accuracy, and calibration error on CIFAR10, CIFAR100, ImageNet, and their out-of-distribution variants compared to previous methods. | accept |
Title: Step-wise Sensitivity Analysis: Identifying Partially Distributed Representations for Interpretable Deep Learning. Abstract: In this paper, we introduce a novel method, called step-wise sensitivity analysis, which makes three contributions towards increasing the interpretability of Deep Neural Networks (DNNs). First, we are the first to suggest a methodology that aggregates results across input stimuli to gain model-centric results. Second, we linearly approximate the neuron activation and propose to use the outlier weights to identify distributed code. Third, our method constructs a dependency graph of the relevant neurons across the network to gain fine-grained understanding of the nature and interactions of DNN's internal features. The dependency graph illustrates shared subgraphs that generalise across 10 classes and can be clustered into semantically related groups. This is the first step towards building decision trees as an interpretation of learned representations. | reject |
Title: Avoiding Robust Misclassifications for Improved Robustness without Accuracy Loss. Abstract: While current methods for training robust deep learning models optimize robust accuracy, in practice, the resulting models are often both robust and inaccurate on numerous samples, providing a false sense of safety for those. Further, they significantly reduce natural accuracy, which hinders the adoption in practice. In this work, we address both of these challenges by extending prior works in three main directions. First, we propose a new training method that jointly maximizes robust accuracy and minimizes robust inaccuracy. Second, since the resulting models are trained to be robust only if they are accurate, we leverage robustness as a principled abstain mechanism. Finally, this abstain mechanism allows us to combine models in a compositional architecture that significantly boosts overall robustness without sacrificing accuracy. We demonstrate the effectiveness of our approach to both empirical and certified robustness on six recent state-of-the-art models and using several datasets. Our results show that our method effectively reduces robust and inaccurate samples by up to 97.28%. Further, it successfully enhanced the $\epsilon_\infty = 1/255$ robustness of a state-of-the-art model from 26% to 86% while only marginally reducing its natural accuracy from 97.8% to 97.6%. | reject |
Title: Generative Ensembles for Robust Anomaly Detection. Abstract: Deep generative models are capable of learning probability distributions over large, high-dimensional datasets such as images, video and natural language. Generative models trained on samples from p(x) ought to assign low likelihoods to out-of-distribution (OoD) samples from q(x), making them suitable for anomaly detection applications. We show that in practice, likelihood models are themselves susceptible to OoD errors, and even assign large likelihoods to images from other natural datasets. To mitigate these issues, we propose Generative Ensembles, a model-independent technique for OoD detection that combines density-based anomaly detection with uncertainty estimation. Our method outperforms ODIN and VIB baselines on image datasets, and achieves comparable performance to a classification model on the Kaggle Credit Fraud dataset. | reject |
Title: Importance and Coherence: Methods for Evaluating Modularity in Neural Networks. Abstract: As deep neural networks become more advanced and widely-used, it is important to understand their inner workings. Toward this goal, modular interpretations are appealing because they offer flexible levels of abstraction aside from standard architectural building blocks (e.g., neurons, channels, layers). In this paper, we consider the problem of assessing how functionally interpretable a given partitioning of neurons is. We propose two proxies for this: importance which reflects how crucial sets of neurons are to network performance, and coherence which reflects how consistently their neurons associate with input/output features. To measure these proxies, we develop a set of statistical methods based on techniques that have conventionally been used for the interpretation of individual neurons. We apply these methods on partitionings generated by a spectral clustering algorithm which uses a graph representation of the network's neurons and weights. We show that despite our partitioning algorithm using neither activations nor gradients, it reveals clusters with a surprising amount of importance and coherence. Together, these results support the use of modular interpretations, and graph-based partitionings in particular, for interpretability. | reject |
Title: Generalized Fourier Features for Coordinate-Based Learning of Functions on Manifolds. Abstract: Recently, positional encoding of input coordinates has been found crucial to enable learning of high-frequency functions with multilayer perceptrons taking low-dimensional coordinate values. In this setting, sinusoids are typically used as a basis for the encoding, which is commonly referred to as "Fourier Features". However, using sinusoids as a basis assumes that the input coordinates lie on Euclidean space. In this work, we generalize positional encoding with Fourier features to non-Euclidean manifolds. We find appropriate bases for positional encoding on manifolds through generalizations of Fourier series. By ensuring the encodings lie on a hypersphere and that the appropriate shifts on the manifold preserve inner-products between encodings, our model approximates convolutions on the manifold, according to the neural tangent kernel (NTK) assumptions. We demonstrate our method on various tasks on different manifolds: 1) learning panoramas on the sphere, 2) learning probability distributions on the rotation manifold, 3) learning neural radiance fields on the product of cube and sphere, and 4) learning light fields represented as the product of spheres. | reject |
Title: Composable Semi-parametric Modelling for Long-range Motion Generation. Abstract: Learning diverse and natural behaviors is one of the longstanding goal for creating intelligent characters in the animated world. In this paper, we propose ``COmposable Semi-parametric MOdelling'' (COSMO), a method for generating long range diverse and distinctive behaviors to achieve a specific goal location. Our proposed method learns to model the motion of human by combining the complementary strengths of both non-parametric techniques and parametric ones. Given the starting and ending state, a memory bank is used to retrieve motion references that are provided as source material to a deep network. The synthesis is performed by a deep network that controls the style of the provided motion material and modifies it to become natural. On skeleton datasets with diverse motion, we show that the proposed method outperforms existing parametric and non-parametric baselines. We also demonstrate the generated sequences are useful as subgoals for actual physical execution in the animated world. | reject |
Title: Dynamic Planning Networks. Abstract: We introduce Dynamic Planning Networks (DPN), a novel architecture for deep reinforcement learning, that combines model-based and model-free aspects for online planning. Our architecture learns to dynamically construct plans using a learned state-transition model by selecting and traversing between simulated states and actions to maximize valuable information before acting. In contrast to model-free methods, model-based planning lets the agent efficiently test action hypotheses without performing costly trial-and-error in the environment. DPN learns to efficiently form plans by expanding a single action-conditional state transition at a time instead of exhaustively evaluating each action, reducing the required number of state-transitions during planning by up to 96%. We observe various emergent planning patterns used to solve environments, including classical search methods such as breadth-first and depth-first search. Learning To Plan shows improved data efficiency, performance, and generalization to new and unseen domains in comparison to several baselines. | reject |
Title: PhraseTransformer: Self-Attention using Local Context for Semantic Parsing. Abstract: Semantic parsing is a challenging task whose purpose is to convert a natural language utterance to machine-understandable information representation. Recently, solutions using Neural Machine Translation have achieved many promising results, especially Transformer because of the ability to learn long-range word dependencies. However, the one drawback of adapting the original Transformer to the semantic parsing is the lack of detail in expressing the information of sentences. Therefore, this work proposes a PhraseTransformer architecture that is capable of a more detailed meaning representation by learning the phrase dependencies in the sentence. The main idea is to incorporate Long Short-Term Memory (LSTM) into the Self-Attention mechanism of the original Transformer to capture more local context of phrases. Experimental results show that the proposed model captures the detailed meaning better than Transformer, raises local context awareness and achieves strong competitive performance on Geo, MSParS datasets, and leads to SOTA performance on Atis dataset in methods using Neural Network. | reject |
Title: Bayesian Context Aggregation for Neural Processes. Abstract: Formulating scalable probabilistic regression models with reliable uncertainty estimates has been a long-standing challenge in machine learning research. Recently, casting probabilistic regression as a multi-task learning problem in terms of conditional latent variable (CLV) models such as the Neural Process (NP) has shown promising results. In this paper, we focus on context aggregation, a central component of such architectures, which fuses information from multiple context data points. So far, this aggregation operation has been treated separately from the inference of a latent representation of the target function in CLV models. Our key contribution is to combine these steps into one holistic mechanism by phrasing context aggregation as a Bayesian inference problem. The resulting Bayesian Aggregation (BA) mechanism enables principled handling of task ambiguity, which is key for efficiently processing context information. We demonstrate on a range of challenging experiments that BA consistently improves upon the performance of traditional mean aggregation while remaining computationally efficient and fully compatible with existing NP-based models. | accept |
Title: Generalizing Graph Convolutional Networks via Heat Kernel. Abstract: Graph convolutional networks (GCNs) have emerged as a powerful framework for mining and learning with graphs. A recent study shows that GCNs can be simplified as a linear model by removing nonlinearities and weight matrices across all consecutive layers, resulting the simple graph convolution (SGC) model. In this paper, we aim to understand GCNs and generalize SGC as a linear model via heat kernel (HKGCN), which acts as a low-pass filter on graphs and enables the aggregation of information from extremely large receptive fields. We theoretically show that HKGCN is in nature a continuous propagation model and GCNs without nonlinearities (i.e., SGC) are the discrete versions of it. Its low-pass filter and continuity properties facilitate the fast and smooth convergence of feature propagation. Experiments on million-scale networks show that the linear HKGCN model not only achieves consistently better results than SGC but also can match or even beat advanced GCN models, while maintaining SGC’s superiority in efficiency. | reject |
Title: Learning Neural Random Fields with Inclusive Auxiliary Generators. Abstract: Neural random fields (NRFs), which are defined by using neural networks to implement potential functions in undirected models, provide an interesting family of model spaces for machine learning. In this paper we develop a new approach to learning NRFs with inclusive-divergence minimized auxiliary generator - the inclusive-NRF approach, for continuous data (e.g. images), with solid theoretical examination on exploiting gradient information in model sampling. We show that inclusive-NRFs can be flexibly used in unsupervised/supervised image generation and semi-supervised classification, and empirically to the best of our knowledge, represent the best-performed random fields in these tasks. Particularly, inclusive-NRFs achieve state-of-the-art sample generation quality on CIFAR-10 in both unsupervised and supervised settings. Semi-supervised inclusive-NRFs show strong classification results on par with state-of-the-art generative model based semi-supervised learning methods, and simultaneously achieve superior generation, on the widely benchmarked datasets - MNIST, SVHN and CIFAR-10. | reject |
Title: Measuring Compositional Generalization: A Comprehensive Method on Realistic Data. Abstract: State-of-the-art machine learning methods exhibit limited compositional generalization. At the same time, there is a lack of realistic benchmarks that comprehensively measure this ability, which makes it challenging to find and evaluate improvements. We introduce a novel method to systematically construct such benchmarks by maximizing compound divergence while guaranteeing a small atom divergence between train and test sets, and we quantitatively compare this method to other approaches for creating compositional generalization benchmarks. We present a large and realistic natural language question answering dataset that is constructed according to this method, and we use it to analyze the compositional generalization ability of three machine learning architectures. We find that they fail to generalize compositionally and that there is a surprisingly strong negative correlation between compound divergence and accuracy. We also demonstrate how our method can be used to create new compositionality benchmarks on top of the existing SCAN dataset, which confirms these findings.
| accept |
Title: Regional based query in graph active learning. Abstract: Graph convolution networks (GCN) have emerged as a leading method to classify nodes and graphs. These GCN have been combined with active learning (AL) methods, when a small chosen set of tagged examples can be used. Most AL-GCN use the sample class uncertainty as selection criteria, and not the graph. In contrast, representative sampling uses the graph, but not the prediction. We propose to combine the two and query nodes based on the uncertainty of the graph around them. We here propose two novel methods to select optimal nodes in AL-GCN that explicitly use the graph information to query for optimal nodes. The first method named regional uncertainty is an extension of the classical entropy measure, but instead of sampling nodes with high entropy, we propose to sample nodes surrounded by nodes of different classes, or nodes with high ambiguity. The second method called Adaptive Page-Rank is an extension of the page-rank algorithm, where nodes that have a low probability of being reached by random walks from tagged nodes are selected. We show that the latter is optimal when the fraction of tagged nodes is low, and when this fraction grows to one over the average degree, the regional uncertainty performs better than all existing methods. While we have tested these methods on graphs, such methods can be extended to any classification problem, where a distance can be defined between the input samples. | reject |
Title: Deconstructing the Inductive Biases of Hamiltonian Neural Networks. Abstract: Physics-inspired neural networks (NNs), such as Hamiltonian or Lagrangian NNs, dramatically outperform other learned dynamics models by leveraging strong inductive biases. These models, however, are challenging to apply to many real world systems, such as those that don’t conserve energy or contain contacts, a common setting for robotics and reinforcement learning. In this paper, we examine the inductive biases that make physics-inspired models successful in practice. We show that, contrary to conventional wisdom, the improved generalization of HNNs is the result of modeling acceleration directly and avoiding artificial complexity from the coordinate system, rather than symplectic structure or energy conservation. We show that by relaxing the inductive biases of these models, we can match or exceed performance on energy-conserving systems while dramatically improving performance on practical, non-conservative systems. We extend this approach to constructing transition models for common Mujoco environments, showing that our model can appropriately balance inductive biases with the flexibility required for model-based control. | accept |
Title: Training By Vanilla SGD with Larger Learning Rates. Abstract: The stochastic gradient descent (SGD) method, first proposed in 1950's, has been the foundation for deep-neural-network (DNN) training with numerous enhancements including adding a momentum or adaptively selecting learning rates, or using both strategies and more. A common view for SGD is that the learning rate should be eventually made small in order to reach sufficiently good approximate solutions. Another widely held view is that the vanilla SGD is out of fashion in comparison to many of its modern variations. In this work, we provide a contrarian claim that, when training over-parameterized DNNs, the vanilla SGD can still compete well with, and oftentimes outperform, its more recent variations by simply using learning rates significantly larger than commonly used values. We establish theoretical results to explain this local convergence behavior of SGD on nonconvex functions, and also present computational evidence, across multiple tasks including image classification, speech recognition and natural language processing, to support the practice of using larger learning rates.
| reject |
Title: Responsible Disclosure of Generative Models Using Scalable Fingerprinting. Abstract: Over the past years, deep generative models have achieved a new level of performance. Generated data has become difficult, if not impossible, to be distinguished from real data. While there are plenty of use cases that benefit from this technology, there are also strong concerns on how this new technology can be misused to generate deep fakes and enable misinformation at scale. Unfortunately, current deep fake detection methods are not sustainable, as the gap between real and fake continues to close. In contrast, our work enables a responsible disclosure of such state-of-the-art generative models, that allows model inventors to fingerprint their models, so that the generated samples containing a fingerprint can be accurately detected and attributed to a source. Our technique achieves this by an efficient and scalable ad-hoc generation of a large population of models with distinct fingerprints. Our recommended operation point uses a 128-bit fingerprint which in principle results in more than 10^{38} identifiable models. Experiments show that our method fulfills key properties of a fingerprinting mechanism and achieves effectiveness in deep fake detection and attribution. Code and models are available at https://github.com/ningyu1991/ScalableGANFingerprints. | accept |
Title: Understanding GANs via Generalization Analysis for Disconnected Support. Abstract: This paper provides theoretical analysis of generative adversarial networks (GANs) to explain its advantages over other standard methods of learning probability measures. GANs learn a probability through observations, using the objective function with a generator and a discriminator. While many empirical results indicate that GANs can generate realistic samples, the reason for such successful performance remains unelucidated. This paper focuses the situation where the target probability measure satisfies the disconnected support property, which means a separate support of a probability, and relates it with the advantage of GANs. It is theoretically shown that, unlike other popular models, GANs do not suffer from the decrease of generalization performance caused by the disconnected support property. We rigorously quantify the generalization performance of GANs of a given architecture, and compare it with the performance of the other models. Based on the theory, we also provide a guideline for selecting deep network architecture for GANs. We demonstrate some numerical examples which support our results. | reject |
Title: How to Improve Sample Complexity of SGD over Highly Dependent Data?. Abstract: Conventional machine learning applications typically assume that data samples are independently and identically distributed (i.i.d.). However, many practical scenarios naturally involve a data-generating process that produces highly dependent data samples, which are known to heavily bias the stochastic optimization process and slow down the convergence of learning. In this paper, we conduct a fundamental study on how to facilitate the convergence of SGD over highly dependent data using different popular update schemes. Specifically, with a $\phi$-mixing model that captures both exponential and polynomial decay of the data dependence over time, we show that SGD with periodic data-subsampling achieves an improved sample complexity over the standard SGD in the full spectrum of the $\phi$-mixing data dependence. Moreover, we show that by fully utilizing the data, mini-batch SGD can further substantially improve the sample complexity with highly dependent data. Numerical experiments validate our theory. | reject |
Title: Practical Adversarial Attacks on Brain--Computer Interfaces. Abstract: Deep learning has been widely employed in brain--computer interfaces (BCIs) to decode a subject's intentions based on recorded brain activities enabling direct interaction with computers and machines. BCI systems play a crucial role in motor rehabilitation and have recently experienced a significant market boost as consumer-grade products. Recent studies have shown that deep learning-based BCIs are vulnerable to adversarial attacks. Failures in such systems might cause medical misdiagnoses, physical harm, and financial damages, hence it is of utmost importance to analyze and understand in-depth, potential malicious attacks to develop countermeasures. In this work, we present the first study that analyzes and models adversarial attacks based on physical domain constraints in EEG-based BCIs. Specifically, we assess the robustness of EEGNet which is the current state-of-the-art network for embedded BCIs. We propose new methods to induce denial-of-service attacks and incorporate domain-specific insights and constraints to accomplish two key goals: (i) create smooth adversarial attacks that are physiologically plausible; (ii) consider the realistic case where the attack happens at the origin of the signal acquisition and it propagates on the human head. Our results show that EEGNet is significantly vulnerable to adversarial attacks with an attack success rate of more than 50\%. With our work, we want to raise awareness and incentivize future developments of proper countermeasures. | reject |
Title: DeeperGCN: Training Deeper GCNs with Generalized Aggregation Functions. Abstract: Graph Convolutional Networks (GCNs) have been drawing significant attention with the power of representation learning on graphs. Recent works developed frameworks to train deep GCNs. Such works show impressive results in tasks like point cloud classification and segmentation, and protein interaction prediction. In this work, we study the performance of such deep models in large scale graph datasets from the Open Graph Benchmark (OGB). In particular, we look at the effect of adequately choosing an aggregation function, and its effect on final performance. Common choices of aggregation are mean, max, and sum. It has shown that GCNs are sensitive to such aggregations when applied to different datasets. We further validate this point and propose to alleviate it by introducing a novel Generalized Aggregation Function. Our new aggregation not only covers all commonly used ones, but also can be tuned to learn customized functions for different tasks. Our generalized aggregation is fully differentiable, and thus its parameters can be learned in an end-to-end fashion. We add our generalized aggregation into a deep GCN framework and show it achieves state-of-the-art results in six benchmarks from OGB. | reject |
Title: ACTIVE REFINEMENT OF WEAKLY SUPERVISED MODELS. Abstract: Supervised machine learning (ML) has fueled major advances in several domains such as health, education and governance. However, most modern ML methods rely on vast quantities of point-by-point hand-labeled training data. In domains such as clinical research, where data collection and its careful characterization is particularly expensive and tedious, this reliance on pointillisticaly labeled data is one of the biggest roadblocks to the adoption of modern data-hungry ML algorithms. Data programming, a framework for learning from weak supervision, attempts to overcome this bottleneck by generating probabilistic training labels from simple yet imperfect heuristics obtained a priori from domain experts. We present WARM, Active Refinement of Weakly Supervised Models, a principled approach to iterative and interactive improvement of weakly supervised models via active learning. WARM directs domain experts' attention on a few selected data points that, when annotated, would improve the label model's probabilistic output in terms of accuracy the most. Gradient backpropagation is then used to iteratively update decision parameters of the heuristics of the label model. Experiments on multiple real-world medical classification datasets reveal that WARM can substantially improve the accuracy of probabilistic labels, a direct measure of training data quality, with as few as 30 queries to clinicians. Additional experiments with domain shift and artificial noise in the LFs, demonstrate WARM's ability to adapt heuristics and the end model to changing population characteristics as well as its robustness to mis-specification of domain-expert-acquired LFs. These capabilities make WARM a potentially useful tool for deploying, maintaining, and auditing weakly supervised systems in practice. | reject |
Title: DEBERTA: DECODING-ENHANCED BERT WITH DISENTANGLED ATTENTION. Abstract: Recent progress in pre-trained neural language models has significantly improved the performance of many natural language processing (NLP) tasks. In this paper we propose a new model architecture DeBERTa (Decoding-enhanced BERT with disentangled attention) that improves the BERT and RoBERTa models using two novel techniques. The first is the disentangled attention mechanism, where each word is represented using two vectors that encode its content and position, respectively, and the attention weights among words are computed using disentangled matrices on their contents and relative positions, respectively. Second, an enhanced mask decoder is used to incorporate absolute positions in the decoding layer to predict the masked tokens in model pre-training. In addition, a new virtual adversarial training method is used for fine-tuning to improve models’
generalization. We show that these techniques significantly improve the efficiency of model pre-training and the performance of both natural language understand(NLU) and natural langauge generation (NLG) downstream tasks. Compared to RoBERTa-Large, a DeBERTa model trained on half of the training data performs consistently better on a wide range of NLP tasks, achieving improvements on MNLI by +0.9% (90.2% vs. 91.1%), on SQuAD v2.0 by +2.3% (88.4% vs. 90.7%) and RACE by +3.6% (83.2% vs. 86.8%). Notably, we scale up DeBERTa by training a larger version that consists of 48 Transform layers with 1.5 billion parameters. The significant performance boost makes the single DeBERTa model surpass the human performance on the SuperGLUE benchmark (Wang et al., 2019a) for the first time in terms of macro-average score (89.9 versus 89.8), and the ensemble DeBERTa model sits atop the SuperGLUE leaderboard as of January 6, 2021, outperforming the human baseline by a decent margin (90.3 versus
89.8). The pre-trained DeBERTa models and the source code were released at: https://github.com/microsoft/DeBERTa.
| accept |
Title: Verification of Non-Linear Specifications for Neural Networks. Abstract: Prior work on neural network verification has focused on specifications that are linear functions of the output of the network, e.g., invariance of the classifier output under adversarial perturbations of the input. In this paper, we extend verification algorithms to be able to certify richer properties of neural networks. To do this we introduce the class of convex-relaxable specifications, which constitute nonlinear specifications that can be verified using a convex relaxation. We show that a number of important properties of interest can be modeled within this class, including conservation of energy in a learned dynamics model of a physical system; semantic consistency of a classifier's output labels under adversarial perturbations and bounding errors in a system that predicts the summation of handwritten digits. Our experimental evaluation shows that our method is able to effectively verify these specifications. Moreover, our evaluation exposes the failure modes in models which cannot be verified to satisfy these specifications. Thus, emphasizing the importance of training models not just to fit training data but also to be consistent with specifications. | accept |
Title: GreaseLM: Graph REASoning Enhanced Language Models. Abstract: Answering complex questions about textual narratives requires reasoning over both stated context and the world knowledge that underlies it. However, pretrained language models (LM), the foundation of most modern QA systems, do not robustly represent latent relationships between concepts, which is necessary for reasoning. While knowledge graphs (KG) are often used to augment LMs with structured representations of world knowledge, it remains an open question how to effectively fuse and reason over the KG representations and the language context, which provides situational constraints and nuances. In this work, we propose GreaseLM, a new model that fuses encoded representations from pretrained LMs and graph neural networks over multiple layers of modality interaction operations. Information from both modalities propagates to the other, allowing language context representations to be grounded by structured world knowledge, and allowing linguistic nuances (e.g., negation, hedging) in the context to inform the graph representations of knowledge. Our results on three benchmarks in the commonsense reasoning (i.e., CommonsenseQA, OpenbookQA) and medical question answering (i.e., MedQA-USMLE) domains demonstrate that GreaseLM can more reliably answer questions that require reasoning over both situational constraints and structured knowledge, even outperforming models 8x larger. | accept |
Title: Learning Robust State Abstractions for Hidden-Parameter Block MDPs. Abstract: Many control tasks exhibit similar dynamics that can be modeled as having common latent structure. Hidden-Parameter Markov Decision Processes (HiP-MDPs) explicitly model this structure to improve sample efficiency in multi-task settings.
However, this setting makes strong assumptions on the observability of the state that limit its application in real-world scenarios with rich observation spaces. In this work, we leverage ideas of common structure from the HiP-MDP setting, and extend it to enable robust state abstractions inspired by Block MDPs. We derive instantiations of this new framework for both multi-task reinforcement learning (MTRL) and meta-reinforcement learning (Meta-RL) settings. Further, we provide transfer and generalization bounds based on task and state similarity, along with sample complexity bounds that depend on the aggregate number of samples across tasks, rather than the number of tasks, a significant improvement over prior work. To further demonstrate efficacy of the proposed method, we empirically compare and show improvement over multi-task and meta-reinforcement learning baselines. | accept |
Title: Trust Region Policy Optimisation in Multi-Agent Reinforcement Learning. Abstract: Trust region methods rigorously enabled reinforcement learning (RL) agents to learn monotonically improving policies, leading to superior performance on a variety of tasks. Unfortunately, when it comes to multi-agent reinforcement learning (MARL), the property of monotonic improvement may not simply apply; this is because agents, even in cooperative games, could have conflicting directions of policy updates. As a result, achieving a guaranteed improvement on the joint policy where each agent acts individually remains an open challenge. In this paper, we extend the theory of trust region learning to MARL. Central to our findings are the multi-agent advantage decomposition lemma and the sequential policy update scheme. Based on these, we develop Heterogeneous-Agent Trust Region Policy Optimisation (HATPRO) and Heterogeneous-Agent Proximal Policy Optimisation (HAPPO) algorithms. Unlike many existing MARL algorithms, HATRPO/HAPPO do not need agents to share parameters, nor do they need any restrictive assumptions on decomposibility of the joint value function. Most importantly, we justify in theory the monotonic improvement property of HATRPO/HAPPO. We evaluate the proposed methods on a series of Multi-Agent MuJoCo and StarCraftII tasks. Results show that HATRPO and HAPPO significantly outperform strong baselines such as IPPO, MAPPO and MADDPG on all tested tasks, thereby establishing a new state of the art. | accept |
Title: Learning Exploration Policies for Navigation. Abstract: Numerous past works have tackled the problem of task-driven navigation. But, how to effectively explore a new environment to enable a variety of down-stream tasks has received much less attention. In this work, we study how agents can autonomously explore realistic and complex 3D environments without the context of task-rewards. We propose a learning-based approach and investigate different policy architectures, reward functions, and training paradigms. We find that use of policies with spatial memory that are bootstrapped with imitation learning and finally finetuned with coverage rewards derived purely from on-board sensors can be effective at exploring novel environments. We show that our learned exploration policies can explore better than classical approaches based on geometry alone and generic learning-based exploration techniques. Finally, we also show how such task-agnostic exploration can be used for down-stream tasks. Videos are available at https://sites.google.com/view/exploration-for-nav/. | accept |
Title: Learning with Social Influence through Interior Policy Differentiation. Abstract: Animals develop novel skills not only through the interaction with the environment but also from the influence of the others. In this work we model the social influence into the scheme of reinforcement learning, enabling the agents to learn both from the environment and from their peers. Specifically, we first define a metric to measure the distance between policies then quantitatively derive the definition of uniqueness. Unlike previous precarious joint optimization approaches, the social uniqueness motivation in our work is imposed as a constraint to encourage the agent to learn a policy different from the existing agents while still solve the primal task. The resulting algorithm, namely Interior Policy Differentiation (IPD), brings about performance improvement as well as a collection of policies that solve a given task with distinct behaviors | reject |
Title: Dropout Q-Functions for Doubly Efficient Reinforcement Learning. Abstract: Randomized ensembled double Q-learning (REDQ) (Chen et al., 2021b) has recently achieved state-of-the-art sample efficiency on continuous-action reinforcement learning benchmarks. This superior sample efficiency is made possible by using a large Q-function ensemble. However, REDQ is much less computationally efficient than non-ensemble counterparts such as Soft Actor-Critic (SAC) (Haarnoja et al., 2018a). To make REDQ more computationally efficient, we propose a method of improving computational efficiency called DroQ, which is a variant of REDQ that uses a small ensemble of dropout Q-functions. Our dropout Q-functions are simple Q-functions equipped with dropout connection and layer normalization. Despite its simplicity of implementation, our experimental results indicate that DroQ is doubly (sample and computationally) efficient. It achieved comparable sample efficiency with REDQ, much better computational efficiency than REDQ, and comparable computational efficiency with that of SAC. | accept |
Title: Language Models are Open Knowledge Graphs. Abstract: This paper shows how to construct knowledge graphs (KGs) from pre-trained language models (e.g., BERT, GPT-2/3), without human supervision. Popular KGs (e.g, Wikidata, NELL) are built in either a supervised or semi-supervised manner, requiring humans to create knowledge. Recent deep language models automatically acquire knowledge from large-scale corpora via pre-training. The stored knowledge has enabled the language models to improve downstream NLP tasks, e.g., answering questions, and writing code and articles. In this paper, we propose an unsupervised method to cast the knowledge contained within language models into KGs. We show that KGs are constructed with a single forward pass of the pre-trained language models (without fine-tuning) over the corpora. We demonstrate the quality of the constructed KGs by comparing to two KGs (Wikidata, TAC KBP) created by humans. Our KGs also provide open factual knowledge that is new in the existing KGs. Our code and KGs will be made publicly available. | reject |
Title: A statistical theory of cold posteriors in deep neural networks. Abstract: To get Bayesian neural networks to perform comparably to standard neural networks it is usually necessary to artificially reduce uncertainty using a tempered or cold posterior. This is extremely concerning: if the prior is accurate, Bayes inference/decision theory is optimal, and any artificial changes to the posterior should harm performance. While this suggests that the prior may be at fault, here we argue that in fact, BNNs for image classification use the wrong likelihood. In particular, standard image benchmark datasets such as CIFAR-10 are carefully curated. We develop a generative model describing curation which gives a principled Bayesian account of cold posteriors, because the likelihood under this new generative model closely matches the tempered likelihoods used in past work. | accept |
Title: Pruning Neural Networks at Initialization: Why Are We Missing the Mark?. Abstract: Recent work has explored the possibility of pruning neural networks at initialization. We assess proposals for doing so: SNIP (Lee et al., 2019), GraSP (Wang et al., 2020), SynFlow (Tanaka et al., 2020), and magnitude pruning. Although these methods surpass the trivial baseline of random pruning, they remain below the accuracy of magnitude pruning after training, and we endeavor to understand why. We show that, unlike pruning after training, randomly shuffling the weights these methods prune within each layer or sampling new initial values preserves or improves accuracy. As such, the per-weight pruning decisions made by these methods can be replaced by a per-layer choice of the fraction of weights to prune. This property suggests broader challenges with the underlying pruning heuristics, the desire to prune at initialization, or both. | accept |
Title: Variance Regularizing Adversarial Learning. Abstract: We study how, in generative adversarial networks, variance in the discriminator's output affects the generator's ability to learn the data distribution. In particular, we contrast the results from various well-known techniques for training GANs when the discriminator is near-optimal and updated multiple times per update to the generator. As an alternative, we propose an additional method to train GANs by explicitly modeling the discriminator's output as a bi-modal Gaussian distribution over the real/fake indicator variables. In order to do this, we train the Gaussian classifier to match the target bi-modal distribution implicitly through meta-adversarial training. We observe that our new method, when trained together with a strong discriminator, provides meaningful, non-vanishing gradients. | reject |
Title: Modeling Adversarial Noise for Adversarial Defense. Abstract: Deep neural networks have been demonstrated to be vulnerable to adversarial noise, promoting the development of defense against adversarial attacks. Motivated by the fact that adversarial noise contains well-generalizing features and that the relationship between adversarial data and natural data can help infer natural data and make reliable predictions, in this paper, we study to model adversarial noise by learning the transition relationship between adversarial labels (i.e. the flipped labels used to generate adversarial data) and natural labels (i.e. the ground truth labels of the natural data). Specifically, we introduce an instance-dependent transition matrix to relate adversarial labels and natural labels, which can be seamlessly embedded with the target model (enabling us to model stronger adaptive adversarial noise). Empirical evaluations demonstrate that our method could effectively improve adversarial accuracy. | reject |
Title: Learned Simulators for Turbulence. Abstract: Turbulence simulation with classical numerical solvers requires high-resolution grids to accurately resolve dynamics. Here we train learned simulators at low spatial and temporal resolutions to capture turbulent dynamics generated at high resolution. We show that our proposed model can simulate turbulent dynamics more accurately than classical numerical solvers at the comparably low resolutions across various scientifically relevant metrics. Our model is trained end-to-end from data and is capable of learning a range of challenging chaotic and turbulent dynamics at low resolution, including trajectories generated by the state-of-the-art Athena++ engine. We show that our simpler, general-purpose architecture outperforms various more specialized, turbulence-specific architectures from the learned turbulence simulation literature. In general, we see that learned simulators yield unstable trajectories; however, we show that tuning training noise and temporal downsampling solves this problem. We also find that while generalization beyond the training distribution is a challenge for learned models, training noise, added loss constraints, and dataset augmentation can help. Broadly, we conclude that our learned simulator outperforms traditional solvers run on coarser grids, and emphasize that simple design choices can offer stability and robust generalization. | accept |
Title: Learning Visual Representations for Transfer Learning by Suppressing Texture. Abstract: Recent works have shown that features obtained from supervised training of CNNs may over-emphasize texture rather than encoding high-level information. In self-supervised learning, in particular, texture as a low-level cue may provide shortcuts that prevent the network from learning higher-level representations. To address these problems we propose to use classic methods based on anisotropic diffusion to augment training using images with suppressed texture. This simple method helps retain important edge information and suppress texture at the same time.
We report our observations for fully supervised and self-supervised learning tasks like MoCoV2 and Jigsaw and achieve state-of-the-art results on object detection and image classification with eight diverse datasets.
Our method is particularly effective for transfer learning tasks and we observed improved performance on five standard transfer learning datasets.
The large improvements on the Sketch-ImageNet dataset, DTD dataset and
additional visual analyses of saliency maps suggest that our approach helps in learning better representations that transfer well. | reject |
Title: Conditional Contrastive Learning with Kernel. Abstract: Conditional contrastive learning frameworks consider the conditional sampling procedure that constructs positive or negative data pairs conditioned on specific variables. Fair contrastive learning constructs negative pairs, for example, from the same gender (conditioning on sensitive information), which in turn reduces undesirable information from the learned representations; weakly supervised contrastive learning constructs positive pairs with similar annotative attributes (conditioning on auxiliary information), which in turn are incorporated into the representations. Although conditional contrastive learning enables many applications, the conditional sampling procedure can be challenging if we cannot obtain sufficient data pairs for some values of the conditioning variable. This paper presents Conditional Contrastive Learning with Kernel (CCL-K) that converts existing conditional contrastive objectives into alternative forms that mitigate the insufficient data problem. Instead of sampling data according to the value of the conditioning variable, CCL-K uses the Kernel Conditional Embedding Operator that samples data from all available data and assigns weights to each sampled data given the kernel similarity between the values of the conditioning variable. We conduct experiments using weakly supervised, fair, and hard negatives contrastive learning, showing CCL-K outperforms state-of-the-art baselines.
| accept |
Title: A Broad Dataset is All You Need for One-Shot Object Detection. Abstract: Is it possible to detect arbitrary objects from a single example? A central problem of all existing attempts at one-shot object detection 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 improve generalization from seen to unseen classes from 45% to 89% and improve the state-of-the-art on COCO by 5.4 AP50 (from 22.0 to 27.5).
We verify that the effect is caused by the number of categories and not the number of training samples, 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. | reject |
Title: Synthesising Realistic Calcium Traces of Neuronal Populations Using GAN. Abstract: Calcium imaging has become a powerful and popular technique to monitor the activity of large populations of neurons in vivo. However, for ethical considerations and despite recent technical developments, recordings are still constrained to a limited number of trials and animals. This limits the amount of data available from individual experiments and hinders the development of analysis techniques and models for more realistic sizes of neuronal populations. The ability to artificially synthesize realistic neuronal calcium signals could greatly alleviate this problem by scaling up the number of trials. Here, we propose a Generative Adversarial Network (GAN) model to generate realistic calcium signals as seen in neuronal somata with calcium imaging. To this end, we propose CalciumGAN, a model based on the WaveGAN architecture and train it on calcium fluorescent signals with the Wasserstein distance. We test the model on artificial data with known ground-truth and show that the distribution of the generated signals closely resembles the underlying data distribution. Then, we train the model on real calcium traces recorded from the primary visual cortex of behaving mice and confirm that the deconvolved spike trains match the statistics of the recorded data. Together, these results demonstrate that our model can successfully generate realistic calcium traces, thereby providing the means to augment existing datasets of neuronal activity for enhanced data exploration and modelling. | reject |
Title: Deep Layers as Stochastic Solvers. Abstract: We provide a novel perspective on the forward pass through a block of layers in a deep network. In particular, we show that a forward pass through a standard dropout layer followed by a linear layer and a non-linear activation is equivalent to optimizing a convex objective with a single iteration of a $\tau$-nice Proximal Stochastic Gradient method. We further show that replacing standard Bernoulli dropout with additive dropout is equivalent to optimizing the same convex objective with a variance-reduced proximal method. By expressing both fully-connected and convolutional layers as special cases of a high-order tensor product, we unify the underlying convex optimization problem in the tensor setting and derive a formula for the Lipschitz constant $L$ used to determine the optimal step size of the above proximal methods. We conduct experiments with standard convolutional networks applied to the CIFAR-10 and CIFAR-100 datasets and show that replacing a block of layers with multiple iterations of the corresponding solver, with step size set via $L$, consistently improves classification accuracy. | accept |
Title: Score and Lyrics-Free Singing Voice Generation. Abstract: Generative models for singing voice have been mostly concerned with the task of "singing voice synthesis," i.e., to produce singing voice waveforms given musical scores and text lyrics. In this work, we explore a novel yet challenging alternative: singing voice generation without pre-assigned scores and lyrics, in both training and inference time. In particular, we experiment with three different schemes: 1) free singer, where the model generates singing voices without taking any conditions; 2) accompanied singer, where the model generates singing voices over a waveform of instrumental music; and 3) solo singer, where the model improvises a chord sequence first and then uses that to generate voices. We outline the associated challenges and propose a pipeline to tackle these new tasks. This involves the development of source separation and transcription models for data preparation, adversarial networks for audio generation, and customized metrics for evaluation. | reject |
Title: “Style” Transfer for Musical Audio Using Multiple Time-Frequency Representations. Abstract: Neural Style Transfer has become a popular technique for
generating images of distinct artistic styles using convolutional neural networks. This
recent success in image style transfer has raised the question of
whether similar methods can be leveraged to alter the “style” of musical
audio. In this work, we attempt long time-scale high-quality audio transfer
and texture synthesis in the time-domain that captures harmonic,
rhythmic, and timbral elements related to musical style, using examples that
may have different lengths and musical keys. We demonstrate the ability
to use randomly initialized convolutional neural networks to transfer
these aspects of musical style from one piece onto another using 3
different representations of audio: the log-magnitude of the Short Time
Fourier Transform (STFT), the Mel spectrogram, and the Constant-Q Transform
spectrogram. We propose using these representations as a way of
generating and modifying perceptually significant characteristics of
musical audio content. We demonstrate each representation's
shortcomings and advantages over others by carefully designing
neural network structures that complement the nature of musical audio. Finally, we show that the most
compelling “style” transfer examples make use of an ensemble of these
representations to help capture the varying desired characteristics of
audio signals. | reject |
Title: On Incorporating Inductive Biases into VAEs. Abstract: We explain why directly changing the prior can be a surprisingly ineffective mechanism for incorporating inductive biases into variational auto-encoders (VAEs), and introduce a simple and effective alternative approach: Intermediary Latent Space VAEs (InteL-VAEs). InteL-VAEs use an intermediary set of latent variables to control the stochasticity of the encoding process, before mapping these in turn to the latent representation using a parametric function that encapsulates our desired inductive bias(es). This allows us to impose properties like sparsity or clustering on learned representations, and incorporate human knowledge into the generative model. Whereas changing the prior only indirectly encourages behavior through regularizing the encoder, InteL-VAEs are able to directly enforce desired characteristics. Moreover, they bypass the computation and encoder design issues caused by non-Gaussian priors, while allowing for additional flexibility through training of the parametric mapping function. We show that these advantages, in turn, lead to both better generative models and better representations being learned. | accept |
Title: Structural Language Models for Any-Code Generation. Abstract: We address the problem of Any-Code Generation (AnyGen) - generating code without any restriction on the vocabulary or structure. The state-of-the-art in this problem is the sequence-to-sequence (seq2seq) approach, which treats code as a sequence and does not leverage any structural information. We introduce a new approach to AnyGen that leverages the strict syntax of programming languages to model a code snippet as tree structural language modeling (SLM). SLM estimates the probability of the program's abstract syntax tree (AST) by decomposing it into a product of conditional probabilities over its nodes. We present a neural model that computes these conditional probabilities by considering all AST paths leading to a target node. Unlike previous structural techniques that have severely restricted the kinds of expressions that can be generated, our approach can generate arbitrary expressions in any programming language. Our model significantly outperforms both seq2seq and a variety of existing structured approaches in generating Java and C# code. We make our code, datasets, and models available online. | reject |
Title: Semi-supervised Outlier Detection using Generative And Adversary Framework. Abstract: In a conventional binary/multi-class classification task, the decision boundary is supported by data from two or more classes. However, in one-class classification task, only data from one class are available. To build an robust outlier detector using only data from a positive class, we propose a corrupted GAN(CorGAN), a deep convolutional Generative Adversary Network requiring no convergence during training. In the adversarial process of training CorGAN, the Generator is supposed to generate outlier samples for negative class, and the Discriminator as an one-class classifier is trained to distinguish data from training datasets (i.e. positive class) and generated data from the Generator (i.e. negative class). To improve the performance of the Discriminator (one-class classifier), we also propose a lot of techniques to improve the performance of the model. The proposed model outperforms the traditional method PCA + PSVM and the solution based on Autoencoder. | reject |
Title: A RECURRENT NEURAL CASCADE-BASED MODEL FOR CONTINUOUS-TIME DIFFUSION PROCESS. Abstract: Many works have been proposed in the literature to capture the dynamics of diffusion in networks. While some of them define graphical markovian models to extract temporal relationships between node infections in networks, others consider diffusion episodes as sequences of infections via recurrent neural models. In this paper we propose a model at the crossroads of these two extremes, which embeds the history of diffusion in infected nodes as hidden continuous states. Depending on the trajectory followed by the content before reaching a given node, the distribution of influence probabilities may vary. However, content trajectories are usually hidden in the data, which induces challenging learning problems. We propose a topological recurrent neural model which exhibits good experimental performances for diffusion modelling and prediction. | reject |
Title: Variance Networks: When Expectation Does Not Meet Your Expectations. Abstract: Ordinary stochastic neural networks mostly rely on the expected values of their weights to make predictions, whereas the induced noise is mostly used to capture the uncertainty, prevent overfitting and slightly boost the performance through test-time averaging. In this paper, we introduce variance layers, a different kind of stochastic layers. Each weight of a variance layer follows a zero-mean distribution and is only parameterized by its variance. It means that each object is represented by a zero-mean distribution in the space of the activations. We show that such layers can learn surprisingly well, can serve as an efficient exploration tool in reinforcement learning tasks and provide a decent defense against adversarial attacks. We also show that a number of conventional Bayesian neural networks naturally converge to such zero-mean posteriors. We observe that in these cases such zero-mean parameterization leads to a much better training objective than more flexible conventional parameterizations where the mean is being learned. | accept |
Title: Integrating linguistic knowledge into DNNs: Application to online grooming detection. Abstract: Online grooming (OG) of children is a pervasive issue in an increasingly interconnected world. We explore various complementary methods to incorporate Corpus Linguistics (CL) knowledge into accurate and interpretable Deep Learning (DL) models. They provide an implicit text normalisation that adapts embedding spaces to the groomers' usage of language, and they focus the DNN's attention onto the expressions of OG strategies. We apply these integration to two architecture types and improve on the state-of-the-art on a new OG corpus. | reject |
Title: Model-Based Reinforcement Learning via Latent-Space Collocation. Abstract: The ability to construct and execute long-term plans enables intelligent agents to solve complex multi-step tasks and prevents myopic behavior only seeking the short-term reward. Recent work has achieved significant progress on building agents that can predict and plan from raw visual observations. However, existing visual planning methods still require a densely shaped reward that provides the algorithm with a short-term signal that is always easy to optimize. These algorithms fail when the shaped reward is not available as they use simplistic planning methods such as sampling-based random shooting and are unable to plan for a distant goal. Instead, to achieve long-horizon visual control, we propose to use collocation-based planning, a powerful optimal control technique that plans forward a sequence of states while constraining the transitions to be physical. We propose a planning algorithm that adapts collocation to visual planning by leveraging probabilistic latent variable models. A model-based reinforcement learning agent equipped with our planning algorithm significantly outperforms prior model-based agents on challenging visual control tasks with sparse rewards and long-term goals. | reject |
Title: Stabilizing DARTS with Amended Gradient Estimation on Architectural Parameters. Abstract: Differentiable neural architecture search has been a popular methodology of exploring architectures for deep learning. Despite the great advantage of search efficiency, it often suffers weak stability, which obstacles it from being applied to a large search space or being flexibly adjusted to different scenarios. This paper investigates DARTS, the currently most popular differentiable search algorithm, and points out an important factor of instability, which lies in its approximation on the gradients of architectural parameters. In the current status, the optimization algorithm can converge to another point which results in dramatic inaccuracy in the re-training process. Based on this analysis, we propose an amending term for computing architectural gradients by making use of a direct property of the optimality of network parameter optimization. Our approach mathematically guarantees that gradient estimation follows a roughly correct direction, which leads the search stage to converge on reasonable architectures. In practice, our algorithm is easily implemented and added to DARTS-based approaches efficiently. Experiments on CIFAR and ImageNet demonstrate that our approach enjoys accuracy gain and, more importantly, enables DARTS-based approaches to explore much larger search spaces that have not been studied before. | reject |
Title: GLAD: Learning Sparse Graph Recovery. Abstract: Recovering sparse conditional independence graphs from data is a fundamental problem in machine learning with wide applications. A popular formulation of the problem is an $\ell_1$ regularized maximum likelihood estimation. Many convex optimization algorithms have been designed to solve this formulation to recover the graph structure. Recently, there is a surge of interest to learn algorithms directly based on data, and in this case, learn to map empirical covariance to the sparse precision matrix. However, it is a challenging task in this case, since the symmetric positive definiteness (SPD) and sparsity of the matrix are not easy to enforce in learned algorithms, and a direct mapping from data to precision matrix may contain many parameters. We propose a deep learning architecture, GLAD, which uses an Alternating Minimization (AM) algorithm as our model inductive bias, and learns the model parameters via supervised learning. We show that GLAD learns a very compact and effective model for recovering sparse graphs from data. | accept |
Title: Context Dependent Modulation of Activation Function. Abstract: We propose a modification to traditional Artificial Neural Networks (ANNs), which provides the ANNs with new aptitudes motivated by biological neurons. Biological neurons work far beyond linearly summing up synaptic inputs and then transforming the integrated information. A biological neuron change firing modes accordingly to peripheral factors (e.g., neuromodulators) as well as intrinsic ones. Our modification connects a new type of ANN nodes, which mimic the function of biological neuromodulators and are termed modulators, to enable other traditional ANN nodes to adjust their activation sensitivities in run-time based on their input patterns. In this manner, we enable the slope of the activation function to be context dependent. This modification produces statistically significant improvements in comparison with traditional ANN nodes in the context of Convolutional Neural Networks and Long Short-Term Memory networks. | reject |
Title: Generalizing Tree Models for Improving Prediction Accuracy. Abstract: Can we generalize and improve the representation power of tree models? Tree models are often favored over deep neural networks due to their interpretable structures in problems where the interpretability is required, such as in the classification of feature-based data where each feature is meaningful. However, most tree models have low accuracies and easily overfit to training data. In this work, we propose Decision Transformer Network (DTN), our highly accurate and interpretable tree model based on our generalized framework of tree models, decision transformers. Decision transformers allow us to describe tree models in the context of deep learning. Our DTN is proposed based on improving the generalizable components of the decision transformer, which increases the representation power of tree models while preserving the inherent interpretability of the tree structure. Our extensive experiments on 121 feature-based datasets show that DTN outperforms the state-of-the-art tree models and even deep neural networks. | reject |
Title: Whitening and second order optimization both destroy information about the dataset, and can make generalization impossible. Abstract: Machine learning is predicated on the concept of generalization: a model achieving low error on a sufficiently large training set should also perform well on novel samples from the same distribution. We show that both data whitening and second order optimization can harm or entirely prevent generalization. In general, model training harnesses information contained in the sample-sample second moment matrix of a dataset. For a general class of models, namely models with a fully connected first layer, we prove that the information contained in this matrix is the only information which can be used to generalize. Models trained using whitened data, or with certain second order optimization schemes, have less access to this information; in the high dimensional regime they have no access at all, resulting in poor or nonexistent generalization ability. We experimentally verify these predictions for several architectures, and further demonstrate that generalization continues to be harmed even when theoretical requirements are relaxed. However, we also show experimentally that regularized second order optimization can provide a practical tradeoff, where training is accelerated but less information is lost, and generalization can in some circumstances even improve. | reject |
Title: Mitigating bias in calibration error estimation. Abstract: Building reliable machine learning systems requires that we correctly understand their level of confidence. Calibration focuses on measuring the degree of accuracy in a model's confidence and most research in calibration focuses on techniques to improve an empirical estimate of calibration error, $\mathrm{ECE}_\mathrm{BIN}$. Using simulation, we show that $\mathrm{ECE}_\mathrm{BIN}$ can systematically underestimate or overestimate the true calibration error depending on the nature of model miscalibration, the size of the evaluation data set, and the number of bins. Critically, $\mathrm{ECE}_\mathrm{BIN}$ is more strongly biased for perfectly calibrated models. We propose a simple alternative calibration error metric, $\mathrm{ECE}_\mathrm{SWEEP}$, in which the number of bins is chosen to be as large as possible while preserving monotonicity in the calibration function. Evaluating our measure on distributions fit to neural network confidence scores on CIFAR-10, CIFAR-100, and ImageNet, we show that $\mathrm{ECE}_\mathrm{SWEEP}$ produces a less biased estimator of calibration error and therefore should be used by any researcher wishing to evaluate the calibration of models trained on similar datasets. | reject |
Title: Decoupling Exploration and Exploitation for Meta-Reinforcement Learning without Sacrifices. Abstract: The goal of meta-reinforcement learning (meta-RL) is to build agents that can quickly learn new tasks by leveraging prior experience on related tasks. Learning a new task often requires both exploring to gather task-relevant information and exploiting this information to solve the task. In principle, optimal exploration and exploitation can be learned end-to-end by simply maximizing task performance. However, such meta-RL approaches struggle with local optima due to a chicken-and-egg problem: learning to explore requires good exploitation to gauge the exploration's utility, but learning to exploit requires information gathered via exploration. Optimizing separate objectives for exploration and exploitation can avoid this problem, but prior meta-RL exploration objectives yield suboptimal policies that gather information irrelevant to the task. We alleviate both concerns by constructing an exploitation objective that automatically identifies task-relevant information and an exploration objective to recover only this information. This avoids local optima in end-to-end training, without sacrificing optimal exploration. Empirically, DREAM substantially outperforms existing approaches on complex meta-RL problems, such as sparse-reward 3D visual navigation. | reject |
Title: Non-Sequential Melody Generation. Abstract: In this paper we present a method for algorithmic melody generation using a generative adversarial network without recurrent components. Music generation has been successfully done using recurrent neural networks, where the model learns sequence information that can help create authentic sounding melodies. Here, we use DCGAN architecture with dilated convolutions and towers to capture sequential information as spatial image information, and learn long-range dependencies in fixed-length melody forms such as Irish traditional reel. | reject |
Title: Episodic Reinforcement Learning with Associative Memory. Abstract: Sample efficiency has been one of the major challenges for deep reinforcement learning. Non-parametric episodic control has been proposed to speed up parametric reinforcement learning by rapidly latching on previously successful policies. However, previous work on episodic reinforcement learning neglects the relationship between states and only stored the experiences as unrelated items. To improve sample efficiency of reinforcement learning, we propose a novel framework, called Episodic Reinforcement Learning with Associative Memory (ERLAM), which associates related experience trajectories to enable reasoning effective strategies. We build a graph on top of states in memory based on state transitions and develop a reverse-trajectory propagation strategy to allow rapid value propagation through the graph. We use the non-parametric associative memory as early guidance for a parametric reinforcement learning model. Results on navigation domain and Atari games show our framework achieves significantly higher sample efficiency than state-of-the-art episodic reinforcement learning models. | accept |
Title: MINE: Mutual Information Neural Estimation. Abstract: This paper presents a Mutual Information Neural Estimator (MINE) that is linearly scalable in dimensionality as well as in sample size. MINE is back-propable and we prove that it is strongly consistent. We illustrate a handful of applications in which MINE is succesfully applied to enhance the property of generative models in both unsupervised and supervised settings. We apply our framework to estimate the information bottleneck, and apply it in tasks related to supervised classification problems. Our results demonstrate substantial added flexibility and improvement in these settings.
| reject |
Title: LEARNED STEP SIZE QUANTIZATION. Abstract: Deep networks run with low precision operations at inference time offer power and space advantages over high precision alternatives, but need to overcome the challenge of maintaining high accuracy as precision decreases. Here, we present a method for training such networks, Learned Step Size Quantization, that achieves the highest accuracy to date on the ImageNet dataset when using models, from a variety of architectures, with weights and activations quantized to 2-, 3- or 4-bits of precision, and that can train 3-bit models that reach full precision baseline accuracy. Our approach builds upon existing methods for learning weights in quantized networks by improving how the quantizer itself is configured. Specifically, we introduce a novel means to estimate and scale the task loss gradient at each weight and activation layer's quantizer step size, such that it can be learned in conjunction with other network parameters. This approach works using different levels of precision as needed for a given system and requires only a simple modification of existing training code. | accept |
Title: Meta-Learning and Universality: Deep Representations and Gradient Descent can Approximate any Learning Algorithm. Abstract: Learning to learn is a powerful paradigm for enabling models to learn from data more effectively and efficiently. A popular approach to meta-learning is to train a recurrent model to read in a training dataset as input and output the parameters of a learned model, or output predictions for new test inputs. Alternatively, a more recent approach to meta-learning aims to acquire deep representations that can be effectively fine-tuned, via standard gradient descent, to new tasks. In this paper, we consider the meta-learning problem from the perspective of universality, formalizing the notion of learning algorithm approximation and comparing the expressive power of the aforementioned recurrent models to the more recent approaches that embed gradient descent into the meta-learner. In particular, we seek to answer the following question: does deep representation combined with standard gradient descent have sufficient capacity to approximate any learning algorithm? We find that this is indeed true, and further find, in our experiments, that gradient-based meta-learning consistently leads to learning strategies that generalize more widely compared to those represented by recurrent models. | accept |
Title: Neural networks behave as hash encoders: An empirical study. Abstract: The input space of a neural network with ReLU-like activations is partitioned into multiple linear regions, each corresponding to a specific activation pattern of the included ReLU-like activations. We demonstrate that this partition exhibits the following encoding properties across a variety of deep learning models: (1) {\it determinism}: almost every linear region contains at most one training example. We can therefore represent almost every training example by a unique activation pattern, which is parameterized by a {\it neural code}; and (2) {\it categorization}: according to the neural code, simple algorithms, such as $K$-Means, $K$-NN, and logistic regression, can achieve fairly good performance on both training and test data. These encoding properties surprisingly suggest that {\it normal neural networks well-trained for classification behave as hash encoders without any extra efforts.} In addition, the encoding properties exhibit variability in different scenarios. {Further experiments demonstrate that {\it model size}, {\it training time}, {\it training sample size}, {\it regularization}, and {\it label noise} contribute in shaping the encoding properties, while the impacts of the first three are dominant.} We then define an {\it activation hash phase chart} to represent the space expanded by {model size}, training time, training sample size, and the encoding properties, which is divided into three canonical regions: {\it under-expressive regime}, {\it critically-expressive regime}, and {\it sufficiently-expressive regime}. | reject |
Title: No MCMC for me: Amortized sampling for fast and stable training of energy-based models. Abstract: Energy-Based Models (EBMs) present a flexible and appealing way to represent uncertainty. Despite recent advances, training EBMs on high-dimensional data remains a challenging problem as the state-of-the-art approaches are costly, unstable, and require considerable tuning and domain expertise to apply successfully. In this work, we present a simple method for training EBMs at scale which uses an entropy-regularized generator to amortize the MCMC sampling typically used in EBM training. We improve upon prior MCMC-based entropy regularization methods with a fast variational approximation. We demonstrate the effectiveness of our approach by using it to train tractable likelihood models. Next, we apply our estimator to the recently proposed Joint Energy Model (JEM), where we match the original performance with faster and stable training. This allows us to extend JEM models to semi-supervised classification on tabular data from a variety of continuous domains. | accept |
Title: Why Do Neural Response Generation Models Prefer Universal Replies?. Abstract: Recent advances in neural Sequence-to-Sequence (Seq2Seq) models reveal a purely data-driven approach to the response generation task. Despite its diverse variants and applications, the existing Seq2Seq models are prone to producing short and generic replies, which blocks such neural network architectures from being utilized in practical open-domain response generation tasks. In this research, we analyze this critical issue from the perspective of the optimization goal of models and the specific characteristics of human-to-human conversational corpora. Our analysis is conducted by decomposing the goal of Neural Response Generation (NRG) into the optimizations of word selection and ordering. It can be derived from the decomposing that Seq2Seq based NRG models naturally tend to select common words to compose responses, and ignore the semantic of queries in word ordering. On the basis of the analysis, we propose a max-marginal ranking regularization term to avoid Seq2Seq models from producing the generic and uninformative responses. The empirical experiments on benchmarks with several metrics have validated our analysis and proposed methodology. | reject |
Title: LEARNING TO SHARE: SIMULTANEOUS PARAMETER TYING AND SPARSIFICATION IN DEEP LEARNING. Abstract: Deep neural networks (DNNs) usually contain millions, maybe billions, of parameters/weights, making both storage and computation very expensive. This has motivated a large body of work to reduce the complexity of the neural network by using sparsity-inducing regularizers. Another well-known approach for controlling the complexity of DNNs is parameter sharing/tying, where certain sets of weights are forced to share a common value. Some forms of weight sharing are hard-wired to express certain in- variances, with a notable example being the shift-invariance of convolutional layers. However, there may be other groups of weights that may be tied together during the learning process, thus further re- ducing the complexity of the network. In this paper, we adopt a recently proposed sparsity-inducing regularizer, named GrOWL (group ordered weighted l1), which encourages sparsity and, simulta- neously, learns which groups of parameters should share a common value. GrOWL has been proven effective in linear regression, being able to identify and cope with strongly correlated covariates. Unlike standard sparsity-inducing regularizers (e.g., l1 a.k.a. Lasso), GrOWL not only eliminates unimportant neurons by setting all the corresponding weights to zero, but also explicitly identifies strongly correlated neurons by tying the corresponding weights to a common value. This ability of GrOWL motivates the following two-stage procedure: (i) use GrOWL regularization in the training process to simultaneously identify significant neurons and groups of parameter that should be tied together; (ii) retrain the network, enforcing the structure that was unveiled in the previous phase, i.e., keeping only the significant neurons and enforcing the learned tying structure. We evaluate the proposed approach on several benchmark datasets, showing that it can dramatically compress the network with slight or even no loss on generalization performance.
| accept |
Title: TAda! Temporally-Adaptive Convolutions for Video Understanding. Abstract: Spatial convolutions are widely used in numerous deep video models. It fundamentally assumes spatio-temporal invariance, i.e., using shared weights for every location in different frames. This work presents Temporally-Adaptive Convolutions (TAdaConv) for video understanding, which shows that adaptive weight calibration along the temporal dimension is an efficient way to facilitate modelling complex temporal dynamics in videos. Specifically, TAdaConv empowers the spatial convolutions with temporal modelling abilities by calibrating the convolution weights for each frame according to its local and global temporal context. Compared to previous temporal modelling operations, TAdaConv is more efficient as it operates over the convolution kernels instead of the features, whose dimension is an order of magnitude smaller than the spatial resolutions. Further, the kernel calibration brings an increased model capacity. We construct TAda2D and TAdaConvNeXt networks by replacing the 2D convolutions in ResNet and ConvNeXt with TAdaConv, which leads to at least on par or better performance compared to state-of-the-art approaches on multiple video action recognition and localization benchmarks. We also demonstrate that as a readily plug-in operation with negligible computation overhead, TAdaConv can effectively improve many existing video models with a convincing margin. | accept |
Title: Metagross: Meta Gated Recursive Controller Units for Sequence Modeling. Abstract: This paper proposes Metagross (Meta Gated Recursive Controller), a new neural sequence modeling unit. Our proposed unit is characterized by recursive parameterization of its gating functions, i.e., gating mechanisms of Metagross are controlled by instances of itself, which are repeatedly called in a recursive fashion. This can be interpreted as a form of meta-gating and recursively parameterizing a recurrent model. We postulate that our proposed inductive bias provides modeling benefits pertaining to learning with inherently hierarchically-structured sequence data (e.g., language, logical or music tasks). To this end, we conduct extensive experiments on recursive logic tasks (sorting, tree traversal, logical inference), sequential pixel-by-pixel classification, semantic parsing, code generation, machine translation and polyphonic music modeling, demonstrating the widespread utility of the proposed approach, i.e., achieving state-of-the-art (or close) performance on all tasks. | reject |
Title: Interpreting Deep Classification Models With Bayesian Inference. Abstract: In this paper, we propose a novel approach to interpret a well-trained classification model through systematically investigating effects of its hidden units on prediction making. We search for the core hidden units responsible for predicting inputs as the class of interest under the generative Bayesian inference framework. We model such a process of unit selection as an Indian Buffet Process, and derive a simplified objective function via the MAP asymptotic technique. The induced binary optimization problem is efficiently solved with a continuous relaxation method by attaching a Switch Gate layer to the hidden layers of interest. The resulted interpreter model is thus end-to-end optimized via standard gradient back-propagation. Experiments are conducted with two popular deep convolutional classifiers, respectively well-trained on the MNIST dataset and the CI- FAR10 dataset. The results demonstrate that the proposed interpreter successfully finds the core hidden units most responsible for prediction making. The modified model, only with the selected units activated, can hold correct predictions at a high rate. Besides, this interpreter model is also able to extract the most informative pixels in the images by connecting a Switch Gate layer to the input layer.
| reject |
Title: OUMG: Objective and Universal Metric for Text Generation with Guiding Ability. Abstract: Existing evaluation metrics for text generation rely on comparing candidate sentences to reference sentences. Some text generation tasks, such as story generation and poetry generation, have no fixed optimal answer and cannot match a corresponding reference for each sentence. Therefore, there is a lack of an objective and universal evaluation metric. To this end, we propose OUMG, a general metric that does not depend on reference standards. We train a discriminator to distinguish between human-generated and machine-generated text, which is used to score the sentences generated by the model. These scores reflect how similar the sentences are to human-generated texts. The capability of the discriminator can be measured by its accuracy, so it avoids the subjectivity of human judgments. Furthermore, the trained discriminator can also guide the text generation process to improve model performance. Experiments on poetry generation demonstrate that OUMG can objectively evaluate text generation models without reference standards. After combining the discriminator with the generation model, the original model can produce significantly higher quality results. | reject |
Title: Understanding Domain Randomization for Sim-to-real Transfer. Abstract: Reinforcement learning encounters many challenges when applied directly in the real world. Sim-to-real transfer is widely used to transfer the knowledge learned from simulation to the real world. Domain randomization---one of the most popular algorithms for sim-to-real transfer---has been demonstrated to be effective in various tasks in robotics and autonomous driving. Despite its empirical successes, theoretical understanding on why this simple algorithm works is largely missing. In this paper, we propose a theoretical framework for sim-to-real transfers, in which the simulator is modeled as a set of MDPs with tunable parameters (corresponding to unknown physical parameters such as friction). We provide sharp bounds on the sim-to-real gap---the difference between the value of policy returned by domain randomization and the value of an optimal policy for the real world. We prove that sim-to-real transfer can succeed under mild conditions without any real-world training samples. Our theory also highlights the importance of using memory (i.e., history-dependent policies) in domain randomization. Our proof is based on novel techniques that reduce the problem of bounding the sim-to-real gap to the problem of designing efficient learning algorithms for infinite-horizon MDPs, which we believe are of independent interest. | accept |
Title: Communication in Multi-Agent Reinforcement Learning: Intention Sharing. Abstract: Communication is one of the core components for learning coordinated behavior in multi-agent systems.
In this paper, we propose a new communication scheme named Intention Sharing (IS) for multi-agent reinforcement learning in order to enhance the coordination among agents. In the proposed IS scheme, each agent generates an imagined trajectory by modeling the environment dynamics and other agents' actions. The imagined trajectory is the simulated future trajectory of each agent based on the learned model of the environment dynamics and other agents and represents each agent's future action plan. Each agent compresses this imagined trajectory capturing its future action plan to generate its intention message for communication by applying an attention mechanism to learn the relative importance of the components in the imagined trajectory based on the received message from other agents. Numeral results show that the proposed IS scheme outperforms other communication schemes in multi-agent reinforcement learning. | accept |
Title: Caveats for information bottleneck in deterministic scenarios. Abstract: Information bottleneck (IB) is a method for extracting information from one random variable X that is relevant for predicting another random variable Y. To do so, IB identifies an intermediate "bottleneck" variable T that has low mutual information I(X;T) and high mutual information I(Y;T). The "IB curve" characterizes the set of bottleneck variables that achieve maximal I(Y;T) for a given I(X;T), and is typically explored by maximizing the "IB Lagrangian", I(Y;T) - βI(X;T). In some cases, Y is a deterministic function of X, including many classification problems in supervised learning where the output class Y is a deterministic function of the input X. We demonstrate three caveats when using IB in any situation where Y is a deterministic function of X: (1) the IB curve cannot be recovered by maximizing the IB Lagrangian for different values of β; (2) there are "uninteresting" trivial solutions at all points of the IB curve; and (3) for multi-layer classifiers that achieve low prediction error, different layers cannot exhibit a strict trade-off between compression and prediction, contrary to a recent proposal. We also show that when Y is a small perturbation away from being a deterministic function of X, these three caveats arise in an approximate way. To address problem (1), we propose a functional that, unlike the IB Lagrangian, can recover the IB curve in all cases. We demonstrate the three caveats on the MNIST dataset. | accept |
Title: On the Importance of Firth Bias Reduction in Few-Shot Classification. Abstract: Learning accurate classifiers for novel categories from very few examples, known as few-shot image classification, is a challenging task in statistical machine learning and computer vision. The performance in few-shot classification suffers from the bias in the estimation of classifier parameters; however, an effective underlying bias reduction technique that could alleviate this issue in training few-shot classifiers has been overlooked. In this work, we demonstrate the effectiveness of Firth bias reduction in few-shot classification. Theoretically, Firth bias reduction removes the $O(N^{-1})$ first order term from the small-sample bias of the Maximum Likelihood Estimator. Here we show that the general Firth bias reduction technique simplifies to encouraging uniform class assignment probabilities for multinomial logistic classification, and almost has the same effect in cosine classifiers. We derive an easy-to-implement optimization objective for Firth penalized multinomial logistic and cosine classifiers, which is equivalent to penalizing the cross-entropy loss with a KL-divergence between the predictions and the uniform label distribution. Then, we empirically evaluate that it is consistently effective across the board for few-shot image classification, regardless of (1) the feature representations from different backbones, (2) the number of samples per class, and (3) the number of classes. Furthermore, we demonstrate the effectiveness of Firth bias reduction on cross-domain and imbalanced data settings. Our implementation is available at https://github.com/ehsansaleh/firth_bias_reduction. | accept |
Title: Robust Natural Language Representation Learning for Natural Language Inference by Projecting Superficial Words out. Abstract: In natural language inference, the semantics of some words do not affect the inference. Such information is considered superficial and brings overfitting. How can we represent and discard such superficial information? In this paper, we use first order logic (FOL) - a classic technique from meaning representation language – to explain what information is superficial for a given sentence pair. Such explanation also suggests two inductive biases according to its properties. We proposed a neural network-based approach that utilizes the two inductive biases. We obtain substantial improvements over extensive experiments. | reject |
Title: Adaptive Pseudo-labeling for Quantum Calculations. Abstract: Machine learning models have recently shown promise in predicting molecular quantum chemical properties. However, the path to real-life adoption requires (1) learning under low-resource constraint and (2) out-of-distribution generalization to unseen, structurally diverse molecules. We observe that these two challenges can be alleviated via abundant labels, which are often not the case in quantum chemistry. We hypothesize that pseudo-labeling on vast array of unlabeled molecules can serve as gold-label proxies to greatly expand the training labeled dataset. The challenge in pseudo-labeling is to prevent the bad pseudo-labels from biasing the model. We develop a simple and effective strategy Pseudo that can assign pseudo-labels, detect bad pseud-labels through evidential uncertainty, and then prevent them from biasing the model using adaptive weighting. Empirically, Pseudo improves quantum calculations accuracy across full data, low data and out-of-distribution settings. | reject |
Title: Adaptive Estimators Show Information Compression in Deep Neural Networks. Abstract: To improve how neural networks function it is crucial to understand their learning process. The information bottleneck theory of deep learning proposes that neural networks achieve good generalization by compressing their representations to disregard information that is not relevant to the task. However, empirical evidence for this theory is conflicting, as compression was only observed when networks used saturating activation functions. In contrast, networks with non-saturating activation functions achieved comparable levels of task performance but did not show compression. In this paper we developed more robust mutual information estimation techniques, that adapt to hidden activity of neural networks and produce more sensitive measurements of activations from all functions, especially unbounded functions. Using these adaptive estimation techniques, we explored compression in networks with a range of different activation functions. With two improved methods of estimation, firstly, we show that saturation of the activation function is not required for compression, and the amount of compression varies between different activation functions. We also find that there is a large amount of variation in compression between different network initializations. Secondary, we see that L2 regularization leads to significantly increased compression, while preventing overfitting. Finally, we show that only compression of the last layer is positively correlated with generalization. | accept |
Title: Non-Transferable Learning: A New Approach for Model Ownership Verification and Applicability Authorization. Abstract: As Artificial Intelligence as a Service gains popularity, protecting well-trained models as intellectual property is becoming increasingly important. There are two common types of protection methods: ownership verification and usage authorization. In this paper, we propose Non-Transferable Learning (NTL), a novel approach that captures the exclusive data representation in the learned model and restricts the model generalization ability to certain domains. This approach provides effective solutions to both model verification and authorization. Specifically: 1) For ownership verification, watermarking techniques are commonly used but are often vulnerable to sophisticated watermark removal methods. By comparison, our NTL-based ownership verification provides robust resistance to state-of-the-art watermark removal methods, as shown in extensive experiments with 6 removal approaches over the digits, CIFAR10 & STL10, and VisDA datasets. 2) For usage authorization, prior solutions focus on authorizing specific users to access the model, but authorized users can still apply the model to any data without restriction. Our NTL-based authorization approach instead provides data-centric protection, which we call applicability authorization, by significantly degrading the performance of the model on unauthorized data. Its effectiveness is also shown through experiments on aforementioned datasets. | accept |
Title: MeshMVS: Multi-view Stereo Guided Mesh Reconstruction. Abstract: Deep learning based 3D shape generation methods generally utilize latent features extracted from color images to encode the objects' semantics and guide the shape generation process. These color image semantics only implicitly encode 3D information, potentially limiting the accuracy of the generated shapes. In this paper we propose a multi-view mesh generation method which incorporates geometry information in the color images explicitly by using the features from intermediate 2.5D depth representations of the input images and regularizing the 3D shapes against these depth images. Our system first predicts a coarse 3D volume from the color images by probabilistically merging voxel occupancy grids from individual views. Depth images corresponding to the multi-view color images are predicted which along with the rendered depth images of the coarse shape are used as a contrastive input whose features guide the refinement of the coarse shape through a series of graph convolution networks. Attention-based multi-view feature pooling is proposed to fuse the contrastive depth features from different viewpoints which are fed to the graph convolution networks.
We validate the proposed multi-view mesh generation method on ShapeNet, where we obtain a significant improvement with 34% decrease in chamfer distance to ground truth and 14% increase in the F1-score compared with the state-of-the-art multi-view shape generation method. | reject |
Title: Ada-NETS: Face Clustering via Adaptive Neighbour Discovery in the Structure Space. Abstract: Face clustering has attracted rising research interest recently to take advantage of massive amounts of face images on the web. State-of-the-art performance has been achieved by Graph Convolutional Networks (GCN) due to their powerful representation capacity. However, existing GCN-based methods build face graphs mainly according to $k$NN relations in the feature space, which may lead to a lot of noise edges connecting two faces of different classes. The face features will be polluted when messages pass along these noise edges, thus degrading the performance of GCNs. In this paper, a novel algorithm named Ada-NETS is proposed to cluster faces by constructing clean graphs for GCNs. In Ada-NETS, each face is transformed to a new structure space, obtaining robust features by considering face features of the neighbour images. Then, an adaptive neighbour discovery strategy is proposed to determine a proper number of edges connecting to each face image. It significantly reduces the noise edges while maintaining the good ones to build a graph with clean yet rich edges for GCNs to cluster faces. Experiments on multiple public clustering datasets show that Ada-NETS significantly outperforms current state-of-the-art methods, proving its superiority and generalization. Code is available at https://github.com/damo-cv/Ada-NETS. | accept |
Title: An Investigation into the Role of Author Demographics in ICLR Participation and Review. Abstract: As machine learning conferences grow rapidly, many are concerned that individuals will be left behind on the basis of traits such as gender and geography. We leverage historic ICLR submissions from 2017 to 2021 to investigate the impact of gender and country of origin both on representation and paper review outcomes at ICLR. We also study various hypotheses that could explain gender representation disparities at ICLR, with a focus on factors that impact the likelihood of an author returning to the conference in consecutive years. Finally, we probe the effects of paper topic on the review process and perform a study on how the inclusion of theorems and the number of co-authors impact the success of papers in the review process. | reject |
Title: Laplacian Smoothing Gradient Descent. Abstract: We propose a class of very simple modifications of gradient descent and stochastic gradient descent. We show that when applied to a large variety of machine learning problems, ranging from softmax regression to deep neural nets, the proposed surrogates can dramatically reduce the variance and improve the generalization accuracy. The methods only involve multiplying the usual (stochastic) gradient by the inverse of a positive definitive matrix coming from the discrete Laplacian or its high order generalizations. The theory of Hamilton-Jacobi partial differential equations demonstrates that the implicit version of new algorithm is almost the same as doing gradient descent on a new function which (i) has the same global minima as the original function and (ii) is ``more convex". We show that optimization algorithms with these surrogates converge uniformly in the discrete Sobolev $H_\sigma^p$ sense and reduce the optimality gap for convex optimization problems. We implement our algorithm into both PyTorch and Tensorflow platforms which only involves changing of a few lines of code. The code will be available on Github. | reject |
Title: Stable Weight Decay Regularization. Abstract: Weight decay is a popular regularization technique for training of deep neural networks. Modern deep learning libraries mainly use $L_{2}$ regularization as the default implementation of weight decay. \citet{loshchilov2018decoupled} demonstrated that $L_{2}$ regularization is not identical to weight decay for adaptive gradient methods, such as Adaptive Momentum Estimation (Adam), and proposed Adam with Decoupled Weight Decay (AdamW). However, we found that the popular implementations of weight decay, including $L_{2}$ regularization and decoupled weight decay, in modern deep learning libraries usually damage performance. First, the $L_{2}$ regularization is unstable weight decay for all optimizers that use Momentum, such as stochastic gradient descent (SGD). Second, decoupled weight decay is highly unstable for all adaptive gradient methods. We further propose the Stable Weight Decay (SWD) method to fix the unstable weight decay problem from a dynamical perspective. The proposed SWD method makes significant improvements over $L_{2}$ regularization and decoupled weight decay in our experiments. Simply fixing weight decay in Adam by SWD, with no extra hyperparameter, can outperform complex Adam variants, which have more hyperparameters. | reject |
Title: Transferring Inductive Biases through Knowledge Distillation. Abstract: Having the right inductive biases can be crucial in many tasks or scenarios where data or computing resources are a limiting factor, or where training data is not perfectly representative of the conditions at test time. However, defining, designing, and efficiently adapting inductive biases is not necessarily straightforward. Inductive biases of a model affect its generalisation behaviour and influence the solution it converges to from different aspects. In this paper, we investigate the power of knowledge distillation in transferring the effects of inductive biases of a teacher model to a student model, when they have different architectures.
We consider different families of models: LSTMs vs. Transformers and CNNs vs. MLPs, in the context of tasks and scenarios with linguistics and vision applications, where having the right inductive biases is critical. We train our models in different setups: no knowledge distillation, self-distillation, and distillation using a teacher with a better inductive bias for the task at hand. We show that in the later setup, compared to no distillation and self-distillation, we can not only improve the performance of the students, but also the solutions they converge become similar to their teachers with respect to a wide range of properties, including different task-specific performance metrics, per sample behaviour of the models, representational similarity and how the representational space of the models evolve during training, performance on out-of-distribution datasets, confidence calibration, and finally whether the converged solutions fall within the same basins of attractions. | reject |
Title: Not-So-Random Features. Abstract: We propose a principled method for kernel learning, which relies on a Fourier-analytic characterization of translation-invariant or rotation-invariant kernels. Our method produces a sequence of feature maps, iteratively refining the SVM margin. We provide rigorous guarantees for optimality and generalization, interpreting our algorithm as online equilibrium-finding dynamics in a certain two-player min-max game. Evaluations on synthetic and real-world datasets demonstrate scalability and consistent improvements over related random features-based methods. | accept |
Title: A Reduction-Based Framework for Conservative Bandits and Reinforcement Learning. Abstract: We study bandits and reinforcement learning (RL) subject to a conservative constraint where the agent is asked to perform at least as well as a given baseline policy. This setting is particular relevant in real-world domains including digital marketing, healthcare, production, finance, etc. In this paper, we present a reduction-based framework for conservative bandits and RL, in which our core technique is to calculate the necessary and sufficient budget obtained from running the baseline policy. For lower bounds, we improve the existing lower bound for conservative multi-armed bandits and obtain new lower bounds for conservative linear bandits, tabular RL and low-rank MDP, through a black-box reduction that turns a certain lower bound in the nonconservative setting into a new lower bound in the conservative setting. For upper bounds, in multi-armed bandits, linear bandits and tabular RL, our new upper bounds tighten or match existing ones with significantly simpler analyses. We also obtain a new upper bound for conservative low-rank MDP. | accept |
Title: Robust Pruning at Initialization. Abstract: Overparameterized Neural Networks (NN) display state-of-the-art performance. However, there is a growing need for smaller, energy-efficient, neural networks to be able to use machine learning applications on devices with limited computational resources. A popular approach consists of using pruning techniques. While these techniques have traditionally focused on pruning pre-trained NN (LeCun et al.,1990; Hassibi et al., 1993), recent work by Lee et al. (2018) has shown promising results when pruning at initialization. However, for Deep NNs, such procedures remain unsatisfactory as the resulting pruned networks can be difficult to train and, for instance, they do not prevent one layer from being fully pruned. In this paper, we provide a comprehensive theoretical analysis of Magnitude and Gradient based pruning at initialization and training of sparse architectures. This allows us to propose novel principled approaches which we validate experimentally on a variety of NN architectures. | accept |
Title: Using Deep Siamese Neural Networks to Speed up Natural Products Research. Abstract: Natural products (NPs, compounds derived from plants and animals) are an important source of novel disease treatments. A bottleneck in the search for new NPs is structure determination. One method is to use 2D Nuclear Magnetic Resonance (NMR) imaging, which indicates bonds between nuclei in the compound, and hence is the "fingerprint" of the compound. Computing a similarity score between 2D NMR spectra for a novel compound and a compound whose structure is known helps determine the structure of the novel compound. Standard approaches to this problem do not appear to scale to larger databases of compounds. Here we use deep convolutional Siamese networks to map NMR spectra to a cluster space, where similarity is given by the distance in the space. This approach results in an AUC score that is more than four times better than an approach using Latent Dirichlet Allocation. | reject |
Title: Link Prediction in Hypergraphs using Graph Convolutional Networks. Abstract: Link prediction in simple graphs is a fundamental problem in which new links between nodes are predicted based on the observed structure of the graph. However, in many real-world applications, there is a need to model relationships among nodes which go beyond pairwise associations. For example, in a chemical reaction, relationship among the reactants and products is inherently higher-order. Additionally, there is need to represent the direction from reactants to products. Hypergraphs provide a natural way to represent such complex higher-order relationships. Even though Graph Convolutional Networks (GCN) have recently emerged as a powerful deep learning-based approach for link prediction over simple graphs, their suitability for link prediction in hypergraphs is unexplored -- we fill this gap in this paper and propose Neural Hyperlink Predictor (NHP). NHP adapts GCNs for link prediction in hypergraphs. We propose two variants of NHP --NHP-U and NHP-D -- for link prediction over undirected and directed hypergraphs, respectively. To the best of our knowledge, NHP-D is the first method for link prediction over directed hypergraphs. Through extensive experiments on multiple real-world datasets, we show NHP's effectiveness. | reject |
Title: Learning Joint Wasserstein Auto-Encoders for Joint Distribution Matching. Abstract: We study the joint distribution matching problem which aims at learning bidirectional mappings to match the joint distribution of two domains. This problem occurs in unsupervised image-to-image translation and video-to-video synthesis tasks, which, however, has two critical challenges: (i) it is difficult to exploit sufficient information from the joint distribution; (ii) how to theoretically and experimentally evaluate the generalization performance remains an open question. To address the above challenges, we propose a new optimization problem and design a novel Joint Wasserstein Auto-Encoders (JWAE) to minimize the Wasserstein distance of the joint distributions in two domains. We theoretically prove that the generalization ability of the proposed method can be guaranteed by minimizing the Wasserstein distance of joint distributions. To verify the generalization ability, we apply our method to unsupervised video-to-video synthesis by performing video frame interpolation and producing visually smooth videos in two domains, simultaneously. Both qualitative and quantitative comparisons demonstrate the superiority of our method over several state-of-the-arts. | reject |
Title: Towards trustworthy predictions from deep neural networks with fast adversarial calibration. Abstract: To facilitate a wide-spread acceptance of AI systems guiding decision making in real-world applications, trustworthiness of deployed models is key. That is, it is crucial for predictive models to be uncertainty-aware and yield well-calibrated (and thus trustworthy) predictions for both in-domain samples as well as under domain shift. Recent efforts to account for predictive uncertainty include post-processing steps for trained neural networks, Bayesian neural networks as well as alternative non-Bayesian approaches such as ensemble approaches and evidential deep learning. Here, we propose an efficient yet general modelling approach for obtaining well-calibrated, trustworthy probabilities for samples obtained after a domain shift. We introduce a new training strategy combining an entropy-encouraging loss term with an adversarial calibration loss term and demonstrate that this results in well-calibrated and technically trustworthy predictions for a wide range of perturbations. We comprehensively evaluate previously proposed approaches on different data modalities, a large range of data sets, network architectures and perturbation strategies and observe that our modelling approach substantially outperforms existing state-of-the-art approaches, yielding well-calibrated predictions for both in-domain and out-of domain samples. | reject |
Title: Style-based Encoder Pre-training for Multi-modal Image Synthesis. Abstract: Image-to-image (I2I) translation aims to translate images from one domain to another. To tackle the multi-modal version of I2I translation, where input and output domains have a one-to-many relation, an extra latent input is provided to the generator to specify a particular output. Recent works propose involved training objectives to learn a latent embedding, jointly with the generator, that models the distribution of possible outputs. Alternatively, we study a simple, yet powerful pre-training strategy for multi-modal I2I translation. We first pre-train an encoder, using a proxy task, to encode the style of an image, such as color and texture, into a low-dimensional latent style vector. Then we train a generator to transform an input image along with a style-code to the output domain. Our generator achieves state-of-the-art results on several benchmarks with a training objective that includes just a GAN loss and a reconstruction loss, which simplifies and speeds up the training significantly compared to competing approaches. We further study the contribution of different loss terms to learning the task of multi-modal I2I translation, and finally we show that the learned style embedding is not dependent on the target domain and generalizes well to other domains. | reject |
Title: Teamwork makes von Neumann work:Min-Max Optimization in Two-Team Zero-Sum Games. Abstract: Motivated by recent advances in both theoretical and applied aspects of multiplayer games, spanning from e-sports to multi-agent generative adversarial networks, we focus on min-max optimization in team zero-sum games. In this class of games, players are split in two teams with payoffs equal within the same team and of opposite sign across the opponent team. Unlike the textbook two player zero-sum games, finding a Nash equilibrium in our class can be shown to be $\textsf{CLS}$-hard, i.e., it is unlikely to have a polynomial time algorithm for computing Nash equilibria. Moreover In this generalized framework, we establish that even asymptotic last iterate or time average convergence to a Nash Equilibrium is not possible using Gradient Descent Ascent (GDA), its optimistic variant and extra gradient. Specifically, we present a family of team games whose induced utility is non-multilinear with non-attractive $\textit{per-se}$ mixed Nash Equilibria, as strict saddle points of the underlying optimization landscape. Leveraging techniques from control theory, we complement these negative results by designing a modified GDA that converges locally to Nash equilibria. Finally, we discuss connections of our framework with AI architectures with team competition structure like multi-agent generative adversarial networks. | reject |
Title: Causal Reasoning from Meta-reinforcement learning. Abstract: Discovering and exploiting the causal structure in the environment is a crucial challenge for intelligent agents. Here we explore whether modern deep reinforcement learning can be used to train agents to perform causal reasoning. We adopt a meta-learning approach, where the agent learns a policy for conducting experiments via causal interventions, in order to support a subsequent task which rewards making accurate causal inferences.We also found the agent could make sophisticated counterfactual predictions, as well as learn to draw causal inferences from purely observational data. Though powerful formalisms for causal reasoning have been developed, applying them in real-world domains can be difficult because fitting to large amounts of high dimensional data often requires making idealized assumptions. Our results suggest that causal reasoning in complex settings may benefit from powerful learning-based approaches. More generally, this work may offer new strategies for structured exploration in reinforcement learning, by providing agents with the ability to perform—and interpret—experiments. | reject |
Title: Learning to Encode Text as Human-Readable Summaries using Generative Adversarial Networks. Abstract: Auto-encoders compress input data into a latent-space representation and reconstruct the original data from the representation. This latent representation is not easily interpreted by humans. In this paper, we propose training an auto-encoder that encodes input text into human-readable sentences. The auto-encoder is composed of a generator and a reconstructor. The generator encodes the input text into a shorter word sequence, and the reconstructor recovers the generator input from the generator output.
To make the generator output human-readable, a discriminator restricts the output of the generator to resemble human-written sentences. By taking the generator output as the summary of the input text, abstractive summarization is achieved without document-summary pairs as training data. Promising results are shown on both English and Chinese corpora. | reject |
Title: Exploring Sentence Vectors Through Automatic Summarization. Abstract: Vector semantics, especially sentence vectors, have recently been used successfully in many areas of natural language processing. However, relatively little work has explored the internal structure and properties of spaces of sentence vectors. In this paper, we will explore the properties of sentence vectors by studying a particular real-world application: Automatic Summarization. In particular, we show that cosine similarity between sentence vectors and document vectors is strongly correlated with sentence importance and that vector semantics can identify and correct gaps between the sentences chosen so far and the document. In addition, we identify specific dimensions which are linked to effective summaries. To our knowledge, this is the first time specific dimensions of sentence embeddings have been connected to sentence properties. We also compare the features of different methods of sentence embeddings. Many of these insights have applications in uses of sentence embeddings far beyond summarization. | reject |
Title: Data augmentation instead of explicit regularization. Abstract: Modern deep artificial neural networks have achieved impressive results through models with very large capacity---compared to the number of training examples---that control overfitting with the help of different forms of regularization. Regularization can be implicit, as is the case of stochastic gradient descent or parameter sharing in convolutional layers, or explicit. Most common explicit regularization techniques, such as dropout and weight decay, reduce the effective capacity of the model and typically require the use of deeper and wider architectures to compensate for the reduced capacity. Although these techniques have been proven successful in terms of results, they seem to waste capacity. In contrast, data augmentation techniques reduce the generalization error by increasing the number of training examples and without reducing the effective capacity. In this paper we systematically analyze the effect of data augmentation on some popular architectures and conclude that data augmentation alone---without any other explicit regularization techniques---can achieve the same performance or higher as regularized models, especially when training with fewer examples. | reject |
Title: FedMix: Approximation of Mixup under Mean Augmented Federated Learning. Abstract: Federated learning (FL) allows edge devices to collectively learn a model without directly sharing data within each device, thus preserving privacy and eliminating the need to store data globally. While there are promising results under the assumption of independent and identically distributed (iid) local data, current state-of-the-art algorithms suffer a performance degradation as the heterogeneity of local data across clients increases. To resolve this issue, we propose a simple framework, \emph{Mean Augmented Federated Learning (MAFL)}, where clients send and receive \emph{averaged} local data, subject to the privacy requirements of target applications. Under our framework, we propose a new augmentation algorithm, named \emph{FedMix}, which is inspired by a phenomenal yet simple data augmentation method, Mixup, but does not require local raw data to be directly shared among devices. Our method shows greatly improved performance in the standard benchmark datasets of FL, under highly non-iid federated settings, compared to conventional algorithms. | accept |