sentence
stringlengths 373
5.09k
| label
stringclasses 2
values |
|---|---|
Title: Efficient Inference of Flexible Interaction in Spiking-neuron Networks. Abstract: Hawkes process provides an effective statistical framework for analyzing the time-dependent interaction of neuronal spiking activities. Although utilized in many real applications, the classic Hawkes process is incapable of modelling inhibitory interactions among neurons. Instead, the nonlinear Hawkes process allows for a more flexible influence pattern with excitatory or inhibitory interactions. In this paper, three sets of auxiliary latent variables (Polya-Gamma variables, latent marked Poisson processes and sparsity variables) are augmented to make functional connection weights in a Gaussian form, which allows for a simple iterative algorithm with analytical updates. As a result, an efficient expectation-maximization (EM) algorithm is derived to obtain the maximum a posteriori (MAP) estimate. We demonstrate the accuracy and efficiency performance of our algorithm on synthetic and real data. For real neural recordings, we show our algorithm can estimate the temporal dynamics of interaction and reveal the interpretable functional connectivity underlying neural spike trains. | accept |
Title: Plug-In Inversion: Model-Agnostic Inversion for Vision with Data Augmentations. Abstract: Existing techniques for model inversion typically rely on hard-to-tune regularizers, such as total variation or feature regularization, which must be individually calibrated for each network in order to produce adequate images. In this work, we introduce Plug-In Inversion, which relies on a simple set of augmentations and does not require excessive hyper-parameter tuning. Under our proposed augmentation-based scheme, the same set of augmentation hyper-parameters can be used for inverting a wide range of image classification models, regardless of input dimensions or the architecture. We illustrate the practicality of our approach by inverting Vision Transformers (ViTs) and Multi-Layer Perceptrons (MLPs) trained on the ImageNet dataset, tasks which to the best of our knowledge have not been successfully accomplished by any previous works. | reject |
Title: Pipelined Training with Stale Weights of Deep Convolutional Neural Networks. Abstract: The growth in the complexity of Convolutional Neural Networks (CNNs) is increasing interest in partitioning a network across multiple accelerators during training and pipelining the backpropagation computations over the accelerators. Existing approaches avoid or limit the use of stale weights through techniques such as micro-batching or weight stashing. These techniques either underutilize of accelerators or increase memory footprint. We explore the impact of stale weights on the statistical efficiency and performance in a pipelined backpropagation scheme that maximizes accelerator utilization and keeps memory overhead modest. We use 4 CNNs (LeNet-5, AlexNet, VGG and ResNet) and show that when pipelining is limited to early layers in a network, training with stale weights converges and results in models with comparable inference accuracies to those resulting from non-pipelined training on MNIST and CIFAR-10 datasets; a drop in accuracy of 0.4%, 4%, 0.83% and 1.45% for the 4 networks, respectively. However, when pipelining is deeper in the network, inference accuracies drop significantly. We propose combining pipelined and non-pipelined training in a hybrid scheme to address this drop. We demonstrate the implementation and performance of our pipelined backpropagation in PyTorch on 2 GPUs using ResNet, achieving speedups of up to 1.8X over a 1-GPU baseline, with a small drop in inference accuracy. | reject |
Title: Batch-shaping for learning conditional channel gated networks. Abstract: We present a method that trains large capacity neural networks with significantly improved accuracy and lower dynamic computational cost. This is achieved by gating the deep-learning architecture on a fine-grained-level. Individual convolutional maps are turned on/off conditionally on features in the network. To achieve this, we introduce a new residual block architecture that gates convolutional channels in a fine-grained manner. We also introduce a generally applicable tool batch-shaping that matches the marginal aggregate posteriors of features in a neural network to a pre-specified prior distribution. We use this novel technique to force gates to be more conditional on the data. We present results on CIFAR-10 and ImageNet datasets for image classification, and Cityscapes for semantic segmentation. Our results show that our method can slim down large architectures conditionally, such that the average computational cost on the data is on par with a smaller architecture, but with higher accuracy. In particular, on ImageNet, our ResNet50 and ResNet34 gated networks obtain 74.60% and 72.55% top-1 accuracy compared to the 69.76% accuracy of the baseline ResNet18 model, for similar complexity. We also show that the resulting networks automatically learn to use more features for difficult examples and fewer features for simple examples. | accept |
Title: Learning without Forgetting: Task Aware Multitask Learning for Multi-Modality Tasks. Abstract: Existing joint learning strategies like multi-task or meta-learning focus more on shared learning and have little to no scope for task-specific learning. This creates the need for a distinct shared pretraining phase and a task-specific finetuning phase. The fine-tuning phase creates separate systems for each task, where improving the performance of a particular task necessitates forgetting some of the knowledge garnered in other tasks. Humans, on the other hand, perform task-specific learning in synergy with general domain-based learning. Inspired by these learning patterns in humans, we suggest a simple yet generic task aware framework to incorporate into existing joint learning strategies. The proposed framework computes task-specific representations to modulate model parameters during joint learning. Hence, it performs both shared and task-specific learning in a single-phase resulting in a single model for all the tasks. The single model itself achieves significant performance gains over the existing joint learning strategies. For example, we train a model on Speech Translation (ST), Automatic Speech Recognition (ASR), and Machine Translation (MT) tasks using the proposed task aware joint learning approach. This single model achieves a performance of 28.64 BLEU score on ST MuST-C English-German, WER of 11.61 on ASR TEDLium v3, and BLEU score of 23.35 on MT WMT14 English-German tasks. This sets a new state-of-the-art performance (SOTA) on the ST task while outperforming the existing end-to-end ASR systems with a competitive performance on the MT task. | reject |
Title: Probabilistic Binary Neural Networks. Abstract: Low bit-width weights and activations are an effective way of combating the increasing need for both memory and compute power of Deep Neural Networks. In this work, we present a probabilistic training method for Neural Network with both binary weights and activations, called PBNet. By embracing stochasticity during training, we circumvent the need to approximate the gradient of functions for which the derivative is zero almost always, such as $\textrm{sign}(\cdot)$, while still obtaining a fully Binary Neural Network at test time. Moreover, it allows for anytime ensemble predictions for improved performance and uncertainty estimates by sampling from the weight distribution. Since all operations in a layer of the PBNet operate on random variables, we introduce stochastic versions of Batch Normalization and max pooling, which transfer well to a deterministic network at test time. We evaluate two related training methods for the PBNet: one in which activation distributions are propagated throughout the network, and one in which binary activations are sampled in each layer. Our experiments indicate that sampling the binary activations is an important element for stochastic training of binary Neural Networks.
| reject |
Title: A Mean-Field Theory for Kernel Alignment with Random Features in Generative Adverserial Networks. Abstract: We propose a novel supervised learning method to optimize the kernel in maximum mean discrepancy generative adversarial networks (MMD GANs). Specifically, we characterize a distributionally robust optimization problem to compute a good distribution for the random feature model of Rahimi and Recht to approximate a good kernel function. Due to the fact that the distributional optimization is infinite dimensional, we consider a Monte-Carlo sample average approximation (SAA) to obtain a more tractable finite dimensional optimization problem. We subsequently leverage a particle stochastic gradient descent (SGD) method to solve finite dimensional optimization problems. Based on a mean-field analysis, we then prove that the empirical distribution of the interactive particles system at each iteration of the SGD follows the path of the gradient descent flow on the Wasserstein manifold. We also establish the non-asymptotic consistency of the finite sample estimator. Our empirical evaluation on synthetic data-set as well as MNIST and CIFAR-10 benchmark data-sets indicates that our proposed MMD GAN model with kernel learning indeed attains higher inception scores well as Fr\`{e}chet inception distances and generates better images compared to the generative moment matching network (GMMN) and MMD GAN with untrained kernels. | reject |
Title: Avoiding Catastrophic States with Intrinsic Fear. Abstract: Many practical reinforcement learning problems contain catastrophic states that the optimal policy visits infrequently or never. Even on toy problems, deep reinforcement learners periodically revisit these states, once they are forgotten under a new policy. In this paper, we introduce intrinsic fear, a learned reward shaping that accelerates deep reinforcement learning and guards oscillating policies against periodic catastrophes. Our approach incorporates a second model trained via supervised learning to predict the probability of imminent catastrophe. This score acts as a penalty on the Q-learning objective. Our theoretical analysis demonstrates that the perturbed objective yields the same average return under strong assumptions and an $\epsilon$-close average return under weaker assumptions. Our analysis also shows robustness to classification errors. Equipped with intrinsic fear, our DQNs solve the toy environments and improve on the Atari games Seaquest, Asteroids, and Freeway. | reject |
Title: Information Laundering for Model Privacy. Abstract: In this work, we propose information laundering, a novel framework for enhancing model privacy. Unlike data privacy that concerns the protection of raw data information, model privacy aims to protect an already-learned model that is to be deployed for public use. The private model can be obtained from general learning methods, and its deployment means that it will return a deterministic or random response for a given input query. An information-laundered model consists of probabilistic components that deliberately maneuver the intended input and output for queries of the model, so the model's adversarial acquisition is less likely. Under the proposed framework, we develop an information-theoretic principle to quantify the fundamental tradeoffs between model utility and privacy leakage and derive the optimal design. | accept |
Title: Neural Compositional Denotational Semantics for Question Answering. Abstract: Answering compositional questions requiring multi-step reasoning is challenging for current models. We introduce an end-to-end differentiable model for interpreting questions, which is inspired by formal approaches to semantics. Each span of text is represented by a denotation in a knowledge graph, together with a vector that captures ungrounded aspects of meaning. Learned composition modules recursively combine constituents, culminating in a grounding for the complete sentence which is an answer to the question. For example, to interpret ‘not green’, the model will represent ‘green’ as a set of entities, ‘not’ as a trainable ungrounded vector, and then use this vector to parametrize a composition function to perform a complement operation. For each sentence, we build a parse chart subsuming all possible parses, allowing the model to jointly learn both the composition operators and output structure by gradient descent. We show the model can learn to represent a variety of challenging semantic operators, such as quantifiers, negation, disjunctions and composed relations on a synthetic question answering task. The model also generalizes well to longer sentences than seen in its training data, in contrast to LSTM and RelNet baselines. We will release our code. | reject |
Title: Generalizing Successor Features to continuous domains for Multi-task Learning. Abstract: The deep reinforcement learning (RL) framework has shown great promise to tackle sequential decision-making problems, where the agent learns to behave optimally through interactions with the environment and receiving rewards.
The ability of an RL agent to learn different reward functions concurrently has many benefits, such as the decomposition of task rewards and skill reuse. One obstacle for achieving this, is the amount of data required as well as the capacity of the model for solving multiple tasks. In this paper, we consider the problem of continuous control for various robot manipulation tasks with an explicit representation that promotes skill reuse while learning multiple tasks, related through the reward function. Our approach relies on two key concepts: successor features (SF), a value function representation that decouples the dynamics of the environment from the rewards, and an actor-critic framework that incorporates the learned SF representations.
We propose a practical implementation of successor features in continuous action spaces. We first show how to learn a decomposable representation required by SF. Our proposed methods, is able to learn decoupled state and reward features representations. We study this approach on a non-trivial continuous control problems with compositional structure built into the reward functions of various tasks. | reject |
Title: Neural Task Graph Execution. Abstract: In order to develop a scalable multi-task reinforcement learning (RL) agent that is able to execute many complex tasks, this paper introduces a new RL problem where the agent is required to execute a given task graph which describes a set of subtasks and dependencies among them. Unlike existing approaches which explicitly describe what the agent should do, our problem only describes properties of subtasks and relationships between them, which requires the agent to perform a complex reasoning to find the optimal subtask to execute. To solve this problem, we propose a neural task graph solver (NTS) which encodes the task graph using a recursive neural network. To overcome the difficulty of training, we propose a novel non-parametric gradient-based policy that performs back-propagation over a differentiable form of the task graph to compute the influence of each subtask on the other subtasks. Our NTS is pre-trained to approximate the proposed gradient-based policy and fine-tuned through actor-critic method. The experimental results on a 2D visual domain show that our method to pre-train from the gradient-based policy significantly improves the performance of NTS. We also demonstrate that our agent can perform a complex reasoning to find the optimal way of executing the task graph and generalize well to unseen task graphs. In addition, we compare our agent with a Monte-Carlo Tree Search (MCTS) method showing that our method is much more efficient than MCTS, and the performance of our agent can be further improved by combining with MCTS. The demo video is available at https://youtu.be/e_ZXVS5VutM. | reject |
Title: Differentiable Expectation-Maximization for Set Representation Learning. Abstract: We tackle the set2vec problem, the task of extracting a vector representation from an input set comprised of a variable number of feature vectors. Although recent approaches based on self attention such as (Set)Transformers were very successful due to the capability of capturing complex interaction between set elements, the computational overhead is the well-known downside. The inducing-point attention and the latest optimal transport kernel embedding (OTKE) are promising remedies that attain comparable or better performance with reduced computational cost, by incorporating a fixed number of learnable queries in attention. In this paper we approach the set2vec problem from a completely different perspective. The elements of an input set are considered as i.i.d.~samples from a mixture distribution, and we define our set embedding feed-forward network as the maximum-a-posterior (MAP) estimate of the mixture which is approximately attained by a few Expectation-Maximization (EM) steps. The whole MAP-EM steps are differentiable operations with a fixed number of mixture parameters, allowing efficient auto-diff back-propagation for any given downstream task. Furthermore, the proposed mixture set data fitting framework allows unsupervised set representation learning naturally via marginal likelihood maximization aka the empirical Bayes. Interestingly, we also find that OTKE can be seen as a special case of our framework, specifically a single-step EM with extra balanced assignment constraints on the E-step. Compared to OTKE, our approach provides more flexible set embedding as well as prior-induced model regularization. We evaluate our approach on various tasks demonstrating improved performance over the state-of-the-arts. | accept |
Title: Learning Long-Term Reward Redistribution via Randomized Return Decomposition. Abstract: Many practical applications of reinforcement learning require agents to learn from sparse and delayed rewards. It challenges the ability of agents to attribute their actions to future outcomes. In this paper, we consider the problem formulation of episodic reinforcement learning with trajectory feedback. It refers to an extreme delay of reward signals, in which the agent can only obtain one reward signal at the end of each trajectory. A popular paradigm for this problem setting is learning with a designed auxiliary dense reward function, namely proxy reward, instead of sparse environmental signals. Based on this framework, this paper proposes a novel reward redistribution algorithm, randomized return decomposition (RRD), to learn a proxy reward function for episodic reinforcement learning. We establish a surrogate problem by Monte-Carlo sampling that scales up least-squares-based reward redistribution to long-horizon problems. We analyze our surrogate loss function by connection with existing methods in the literature, which illustrates the algorithmic properties of our approach. In experiments, we extensively evaluate our proposed method on a variety of benchmark tasks with episodic rewards and demonstrate substantial improvement over baseline algorithms. | accept |
Title: A Real-time Contribution Measurement Method for Participants in Federated Learning. Abstract: Federated learning is a framework for protecting distributed data privacy and has participated in commercial activities. However, there is a lack of a sufficiently reasonable contribution measurement mechanism to distribute the reward for each agent. In the commercial union, if there is no mechanism like this, every agent will get the same reward. This is unfair to agents that provide better data, so such a mechanism is needed. To address this issue, this work proposes a real-time contribution measurement method. Firstly, the method defines the impact of each agent. Furthermore, we comprehensively consider the current round and the previous round to obtain the contribution rate of each agent. To verify effectiveness of the proposed method, the work conducts pseudo-distributed training and an experiment on the Penn Treebank dataset. Comparing the Shapley Value in game theory, the comparative experiment result shows that the proposed method is more sensitive to both data quantity and data quality under the premise of maintaining real-time. | reject |
Title: All-You-Can-Fit 8-Bit Flexible Floating-Point Format for Accurate and Memory-Efficient Inference of Deep Neural Networks. Abstract: Modern deep neural network (DNN) models generally require a huge amount of weight and activation values to achieve good inference outcomes. Those data inevitably demand a massive off-chip memory capacity/bandwidth, and the situation gets even worse if they are represented in high-precision floating-point formats. Effort has been made for representing those data in different 8-bit floating-point formats, nevertheless, a notable accuracy loss is still unavoidable. In this paper we introduce an extremely flexible 8-bit floating-point (FFP8) format whose defining factors – the bit width of exponent/fraction field, the exponent bias, and even the presence of the sign bit – are all configurable. We also present a methodology to properly determine those factors so that the accuracy of model inference can be maximized. The foundation of this methodology is based on a key observation – both the maximum magnitude and the value distribution are quite dissimilar between weights and activations in most DNN models. Experimental results demonstrate that the proposed FFP8 format achieves an extremely low accuracy loss of $0.1\%\sim 0.3\%$ for several representative image classification models even without the need of model retraining. Besides, it is easy to turn a classical floating-point processing unit into an FFP8-compliant one, and the extra hardware cost is minor. | reject |
Title: Model agnostic meta-learning on trees. Abstract: In meta-learning, the knowledge learned from previous tasks is transferred to new ones, but this transfer only works if tasks are related, and sharing information between unrelated tasks might hurt performance. A fruitful approach is to share gradients across similar tasks during training, and recent work suggests that the gradients themselves can be used as a measure of task similarity.
We study the case in which datasets associated to different tasks have a hierarchical, tree structure. While a few methods have been proposed for hierarchical meta-learning in the past, we propose the first algorithm that is model-agnostic, a simple extension of MAML. As in MAML, our algorithm adapts the model to each task with a few gradient steps, but the adaptation follows the tree structure: in each step, gradients are pooled across task clusters, and subsequent steps follow down the tree. We test the algorithm on linear and non-linear regression on synthetic data, and show that the algorithm significantly improves over MAML. Interestingly, the algorithm performs best when it does not know in advance the tree structure of the data. | reject |
Title: Deep, Skinny Neural Networks are not Universal Approximators. Abstract: In order to choose a neural network architecture that will be effective for a particular modeling problem, one must understand the limitations imposed by each of the potential options. These limitations are typically described in terms of information theoretic bounds, or by comparing the relative complexity needed to approximate example functions between different architectures. In this paper, we examine the topological constraints that the architecture of a neural network imposes on the level sets of all the functions that it is able to approximate. This approach is novel for both the nature of the limitations and the fact that they are independent of network depth for a broad family of activation functions. | accept |
Title: Self-Contrastive Learning. Abstract: This paper proposes a novel contrastive learning framework, called Self-Contrastive (SelfCon) Learning, that self-contrasts within multiple outputs from the different levels of a multi-exit network. SelfCon learning does not require additional augmented samples, which resolves the concerns of multi-viewed batch (e.g., high computational cost and generalization error). Furthermore, we prove that SelfCon loss guarantees the lower bound of label-conditional mutual information between the intermediate and the last feature. In our experiments including ImageNet-100, SelfCon surpasses cross-entropy and Supervised Contrastive (SupCon) learning without the need for a multi-viewed batch. We demonstrate that the success of SelfCon learning is related to the regularization effect associated with the single-view and sub-network. | reject |
Title: Learning Features with Parameter-Free Layers. Abstract: Trainable layers such as convolutional building blocks are the standard network design choices by learning parameters to capture the global context through successive spatial operations. When designing an efficient network, trainable layers such as the depthwise convolution is the source of efficiency in the number of parameters and FLOPs, but there was little improvement to the model speed in practice. This paper argues that simple built-in parameter-free operations can be a favorable alternative to the efficient trainable layers replacing spatial operations in a network architecture. We aim to break the stereotype of organizing the spatial operations of building blocks into trainable layers. Extensive experimental analyses based on layer-level studies with fully-trained models and neural architecture searches are provided to investigate whether parameter-free operations such as the max-pool are functional. The studies eventually give us a simple yet effective idea for redesigning network architectures, where the parameter-free operations are heavily used as the main building block without sacrificing the model accuracy as much. Experimental results on the ImageNet dataset demonstrate that the network architectures with parameter-free operations could enjoy the advantages of further efficiency in terms of model speed, the number of the parameters, and FLOPs. Code and ImageNet pretrained models are available at https://github.com/naver-ai/PfLayer.
| accept |
Title: Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Bounds. Abstract: We design a new algorithm for batch active learning with deep neural network models. Our algorithm, Batch Active learning by Diverse Gradient Embeddings (BADGE), samples groups of points that are disparate and high-magnitude when represented in a hallucinated gradient space, a strategy designed to incorporate both predictive uncertainty and sample diversity into every selected batch. Crucially, BADGE trades off between diversity and uncertainty without requiring any hand-tuned hyperparameters. While other approaches sometimes succeed for particular batch sizes or architectures, BADGE consistently performs as well or better, making it a useful option for real world active learning problems. | accept |
Title: Self-Supervised State-Control through Intrinsic Mutual Information Rewards. Abstract: Learning to discover useful skills without a manually-designed reward function would have many applications, yet is still a challenge for reinforcement learning. In this paper, we propose Mutual Information-based State-Control (MISC), a new self-supervised Reinforcement Learning approach for learning to control states of interest without any external reward function. We formulate the intrinsic objective as rewarding the skills that maximize the mutual information between the context states and the states of interest. For example, in robotic manipulation tasks, the context states are the robot states and the states of interest are the states of an object. We evaluate our approach for different simulated robotic manipulation tasks from OpenAI Gym. We show that our method is able to learn to manipulate the object, such as pushing and picking up, purely based on the intrinsic mutual information rewards. Furthermore, the pre-trained policy and mutual information discriminator can be used to accelerate learning to achieve high task rewards. Our results show that the mutual information between the context states and the states of interest can be an effective ingredient for overcoming challenges in robotic manipulation tasks with sparse rewards. A video showing experimental results is available at https://youtu.be/cLRrkd3Y7vU | reject |
Title: Learning Self-Imitating Diverse Policies. Abstract: The success of popular algorithms for deep reinforcement learning, such as policy-gradients and Q-learning, relies heavily on the availability of an informative reward signal at each timestep of the sequential decision-making process. When rewards are only sparsely available during an episode, or a rewarding feedback is provided only after episode termination, these algorithms perform sub-optimally due to the difficultly in credit assignment. Alternatively, trajectory-based policy optimization methods, such as cross-entropy method and evolution strategies, do not require per-timestep rewards, but have been found to suffer from high sample complexity by completing forgoing the temporal nature of the problem. Improving the efficiency of RL algorithms in real-world problems with sparse or episodic rewards is therefore a pressing need. In this work, we introduce a self-imitation learning algorithm that exploits and explores well in the sparse and episodic reward settings. We view each policy as a state-action visitation distribution and formulate policy optimization as a divergence minimization problem. We show that with Jensen-Shannon divergence, this divergence minimization problem can be reduced into a policy-gradient algorithm with shaped rewards learned from experience replays. Experimental results indicate that our algorithm works comparable to existing algorithms in environments with dense rewards, and significantly better in environments with sparse and episodic rewards. We then discuss limitations of self-imitation learning, and propose to solve them by using Stein variational policy gradient descent with the Jensen-Shannon kernel to learn multiple diverse policies. We demonstrate its effectiveness on a challenging variant of continuous-control MuJoCo locomotion tasks. | accept |
Title: Meta-learning Transferable Representations with a Single Target Domain. Abstract: Recent works found that fine-tuning and joint training---two popular approaches for transfer learning---do not always improve accuracy on downstream tasks. First, we aim to understand more about when and why fine-tuning and joint training can be suboptimal or even harmful for transfer learning. We design semi-synthetic datasets where the source task can be solved by either source-specific features or transferable features. We observe that (1) pre-training may not have incentive to learn transferable features and (2) joint training may simultaneously learn source-specific features and overfit to the target. Second, to improve over fine-tuning and joint training, we propose Meta Representation Learning MeRLin to learn transferable features. MeRLin meta-learns representations by ensuring that a head fit on top of the representations with target training data also performs well on target validation data. We also prove that MeRLin recovers the target ground-truth model with a quadratic neural net parameterization and a source distribution that contains both transferable and source-specific features. On the same distribution, pre-training and joint training provably fail to learn transferable features. MeRLin empirically outperforms previous state-of-the-art transfer learning algorithms on various real-world vision and NLP transfer learning benchmarks. | reject |
Title: Generalized Energy Based Models. Abstract: We introduce the Generalized Energy Based Model (GEBM) for generative modelling. These models combine two trained components: a base distribution (generally an implicit model), which can learn the support of data with low intrinsic dimension in a high dimensional space; and an energy function, to refine the probability mass on the learned support.
Both the energy function and base jointly constitute the final model, unlike GANs, which retain only the base distribution (the "generator").
GEBMs are trained by alternating between learning the energy and the base.
We show that both training stages are well-defined: the energy is learned by maximising a generalized likelihood, and the resulting energy-based loss provides informative gradients for learning the base.
Samples from the posterior on the latent space of the trained model can be obtained via MCMC, thus finding regions in this space that produce better quality samples.
Empirically, the GEBM samples on image-generation tasks are of much better quality than those from the learned generator alone, indicating that all else being equal, the GEBM will outperform a GAN of the same complexity. When using normalizing flows as base measures, GEBMs succeed on density modelling tasks returning comparable performance to direct maximum likelihood of the same networks. | accept |
Title: Deep Partial Updating. Abstract: Emerging edge intelligence applications require the server to continuously retrain and update deep neural networks deployed on remote edge nodes to leverage newly collected data samples. Unfortunately, it may be impossible in practice to continuously send fully updated weights to these edge nodes due to the highly constrained communication resource. In this paper, we propose the weight-wise deep partial updating paradigm, which smartly selects only a subset of weights to update at each server-to-edge communication round, while achieving a similar performance compared to full updating. Our method is established through analytically upper-bounding the loss difference between partial updating and full updating, and only updates the weights which make the largest contributions to the upper bound. Extensive experimental results demonstrate the efficacy of our partial updating methodology which achieves a high inference accuracy while updating a rather small number of weights. | reject |
Title: Maximum Reward Formulation In Reinforcement Learning. Abstract: Reinforcement learning (RL) algorithms typically deal with maximizing the expected cumulative return (discounted or undiscounted, finite or infinite horizon). However, several crucial applications in the real world, such as drug discovery, do not fit within this framework because an RL agent only needs to identify states (molecules) that achieve the highest reward within a trajectory and does not need to optimize for the expected cumulative return. In this work, we formulate an objective function to maximize the expected maximum reward along a trajectory, derive a novel functional form of the Bellman equation, introduce the corresponding Bellman operators, and provide a proof of convergence. Using this formulation, we achieve state-of-the-art results on the task of molecule generation that mimics a real-world drug discovery pipeline. | reject |
Title: Expressiveness in Deep Reinforcement Learning. Abstract: Representation learning in reinforcement learning (RL) algorithms focuses on extracting useful features for choosing good actions. Expressive representations are essential for learning well-performed policies. In this paper, we study the relationship between the state representation assigned by the state extractor and the performance of the RL agent. We observe that representations assigned by the better state extractor are more scattered than which assigned by the worse one. Moreover, RL agents achieving high performances always have high rank matrices which are composed by their representations. Based on our observations, we formally define expressiveness of the state extractor as the rank of the matrix composed by representations. Therefore, we propose to promote expressiveness so as to improve algorithm performances, and we call it Expressiveness Promoted DRL. We apply our method on both policy gradient and value-based algorithms, and experimental results on 55 Atari games show the superiority of our proposed method. | reject |
Title: Compressed Sensing with Deep Image Prior and Learned Regularization. Abstract: We propose a novel method for compressed sensing recovery using
untrained deep generative models. Our method is based on the recently
proposed Deep Image Prior (DIP), wherein the convolutional weights of
the network are optimized to match the observed measurements. We show
that this approach can be applied to solve any differentiable linear inverse
problem, outperforming previous unlearned methods. Unlike various learned approaches based on generative models, our method does not require pre-training over large datasets. We further introduce a novel learned regularization technique, which incorporates prior information on the network weights. This reduces reconstruction error, especially for noisy measurements. Finally we prove that, using the DIP optimization approach, moderately overparameterized single-layer networks trained can perfectly fit any signal despite the nonconvex nature of the fitting problem. This theoretical result provides justification for early stopping. | reject |
Title: Composable Planning with Attributes. Abstract: The tasks that an agent will need to solve often aren’t known during training. However, if the agent knows which properties of the environment we consider im- portant, then after learning how its actions affect those properties the agent may be able to use this knowledge to solve complex tasks without training specifi- cally for them. Towards this end, we consider a setup in which an environment is augmented with a set of user defined attributes that parameterize the features of interest. We propose a model that learns a policy for transitioning between “nearby” sets of attributes, and maintains a graph of possible transitions. Given a task at test time that can be expressed in terms of a target set of attributes, and a current state, our model infers the attributes of the current state and searches over paths through attribute space to get a high level plan, and then uses its low level policy to execute the plan. We show in grid-world games and 3D block stacking that our model is able to generalize to longer, more complex tasks at test time even when it only sees short, simple tasks at train time.
| reject |
Title: Gap-Aware Mitigation of Gradient Staleness. Abstract: Cloud computing is becoming increasingly popular as a platform for distributed training of deep neural networks. Synchronous stochastic gradient descent (SSGD) suffers from substantial slowdowns due to stragglers if the environment is non-dedicated, as is common in cloud computing. Asynchronous SGD (ASGD) methods are immune to these slowdowns but are scarcely used due to gradient staleness, which encumbers the convergence process. Recent techniques have had limited success mitigating the gradient staleness when scaling up to many workers (computing nodes). In this paper we define the Gap as a measure of gradient staleness and propose Gap-Aware (GA), a novel asynchronous-distributed method that penalizes stale gradients linearly to the Gap and performs well even when scaling to large numbers of workers. Our evaluation on the CIFAR, ImageNet, and WikiText-103 datasets shows that GA outperforms the currently acceptable gradient penalization method, in final test accuracy. We also provide convergence rate proof for GA. Despite prior beliefs, we show that if GA is applied, momentum becomes beneficial in asynchronous environments, even when the number of workers scales up. | accept |
Title: Multi-label Learning for Large Text Corpora using Latent Variable Model with Provable Gurantees. Abstract: Here we study the problem of learning labels for large text corpora where each document can be assigned a variable number of labels. The problem is trivial when the label dimensionality is small and can be easily solved by a series of one-vs-all classifiers. However, as the label dimensionality increases, the parameter space of such one-vs-all classifiers becomes extremely large and outstrips the memory. Here we propose a latent variable model to reduce the size of the parameter space, but still efficiently learn the labels. We learn the model using spectral learning and show how to extract the parameters using only three passes through the training dataset. Further, we analyse the sample complexity of our model using PAC learning theory and then demonstrate the performance of our algorithm on several benchmark datasets in comparison with existing algorithms.
| reject |
Title: DriPP: Driven Point Processes to Model Stimuli Induced Patterns in M/EEG Signals. Abstract: The quantitative analysis of non-invasive electrophysiology signals from electroencephalography (EEG) and magnetoencephalography (MEG) boils down to the identification of temporal patterns such as evoked responses, transient bursts of neural oscillations but also blinks or heartbeats for data cleaning. Several works have shown that these patterns can be extracted efficiently in an unsupervised way, e.g., using Convolutional Dictionary Learning. This leads to an event-based description of the data. Given these events, a natural question is to estimate how their occurrences are modulated by certain cognitive tasks and experimental manipulations. To address it, we propose a point process approach. While point processes have been used in neuroscience in the past, in particular for single cell recordings (spike trains), techniques such as Convolutional Dictionary Learning make them amenable to human studies based on EEG/MEG signals. We develop a novel statistical point process model – called driven temporal point processes (DriPP) – where the intensity function of the point process model is linked to a set of point processes corresponding to stimulation events. We derive a fast and principled expectation-maximization algorithm to estimate the parameters of this model. Simulations reveal that model parameters can be identified from long enough signals. Results on standard MEG datasets demonstrate that our methodology reveals event-related neural responses – both evoked and induced – and isolates non-task specific temporal patterns. | accept |
Title: Discriminator Rejection Sampling. Abstract: We propose a rejection sampling scheme using the discriminator of a GAN to
approximately correct errors in the GAN generator distribution. We show that
under quite strict assumptions, this will allow us to recover the data distribution
exactly. We then examine where those strict assumptions break down and design a
practical algorithm—called Discriminator Rejection Sampling (DRS)—that can be
used on real data-sets. Finally, we demonstrate the efficacy of DRS on a mixture of
Gaussians and on the state of the art SAGAN model. On ImageNet, we train an
improved baseline that increases the best published Inception Score from 52.52 to
62.36 and reduces the Frechet Inception Distance from 18.65 to 14.79. We then use
DRS to further improve on this baseline, improving the Inception Score to 76.08
and the FID to 13.75. | accept |
Title: Monotonic Differentiable Sorting Networks. Abstract: Differentiable sorting algorithms allow training with sorting and ranking supervision, where only the ordering or ranking of samples is known. Various methods have been proposed to address this challenge, ranging from optimal transport-based differentiable Sinkhorn sorting algorithms to making classic sorting networks differentiable. One problem of current differentiable sorting methods is that they are non-monotonic. To address this issue, we propose a novel relaxation of conditional swap operations that guarantees monotonicity in differentiable sorting networks. We introduce a family of sigmoid functions and prove that they produce differentiable sorting networks that are monotonic. Monotonicity ensures that the gradients always have the correct sign, which is an advantage in gradient-based optimization. We demonstrate that monotonic differentiable sorting networks improve upon previous differentiable sorting methods. | accept |
Title: Limitations of Active Learning With Deep Transformer Language Models. Abstract: Active Learning (AL) has the potential to reduce labeling cost when training natural language processing models, but its effectiveness with the large pretrained transformer language models that power today's NLP is uncertain. We present experiments showing that when applied to modern pretrained models, active learning offers inconsistent and often poor performance. As in prior work, we find that AL sometimes selects harmful "unlearnable" collective outliers, but we discover that some failures have a different explanation: the examples AL selects are informative but also increase training instability, reducing average performance. Our findings suggest that for some datasets this instability can be mitigated by training multiple models and selecting the best on a validation set, which we show impacts relative AL performance comparably to the outlier-pruning technique from prior work while also increasing absolute performance. Our experiments span three pretrained models, ten datasets, and four active learning approaches. | reject |
Title: Unconditional Synthesis of Complex Scenes Using a Semantic Bottleneck. Abstract: Coupling the high-fidelity generation capabilities of label-conditional image synthesis methods with the flexibility of unconditional generative models, we propose a semantic bottleneck GAN model for unconditional synthesis of complex scenes. We assume pixel-wise segmentation labels are available during training and use them to learn the scene structure through an unconditional progressive segmentation generation network. During inference, our model first synthesizes a realistic segmentation layout from scratch, then synthesizes a realistic scene conditioned on that layout through a conditional segmentation-to-image synthesis network. When trained end-to-end, the resulting model outperforms state-of-the-art generative models in unsupervised image synthesis on two challenging domains in terms of the Frechet Inception Distance and perceptual evaluations. Moreover, we demonstrate that the end-to-end training significantly improves the segmentation-to-image synthesis sub-network, which results in superior performance over the state-of-the-art when conditioning on real segmentation layouts. | reject |
Title: On the Generalization of Models Trained with SGD: Information-Theoretic Bounds and Implications. Abstract: This paper follows up on a recent work of Neu et al. (2021) and presents some new information-theoretic upper bounds for the generalization error of machine learning models, such as neural networks, trained with SGD. We apply these bounds to analyzing the generalization behaviour of linear and two-layer ReLU networks. Experimental study of these bounds provide some insights on the SGD training of neural networks. They also point to a new and simple regularization scheme which we show performs comparably to the current state of the art. | accept |
Title: Beyond Word Importance: Contextual Decomposition to Extract Interactions from LSTMs. Abstract: The driving force behind the recent success of LSTMs has been their ability to learn complex and non-linear relationships. Consequently, our inability to describe these relationships has led to LSTMs being characterized as black boxes. To this end, we introduce contextual decomposition (CD), an interpretation algorithm for analysing individual predictions made by standard LSTMs, without any changes to the underlying model. By decomposing the output of a LSTM, CD captures the contributions of combinations of words or variables to the final prediction of an LSTM. On the task of sentiment analysis with the Yelp and SST data sets, we show that CD is able to reliably identify words and phrases of contrasting sentiment, and how they are combined to yield the LSTM's final prediction. Using the phrase-level labels in SST, we also demonstrate that CD is able to successfully extract positive and negative negations from an LSTM, something which has not previously been done. | accept |
Title: Pixelated Butterfly: Simple and Efficient Sparse training for Neural Network Models. Abstract: Overparameterized neural networks generalize well but are expensive to train. Ideally one would like to reduce their computational cost while retaining their generalization benefits. Sparse model training is a simple and promising approach to achieve this, but there remain challenges as existing methods struggle with accuracy loss, slow training runtime, or difficulty in sparsifying all model components. The core problem is that searching for a sparsity mask over a discrete set of sparse matrices is difficult and expensive. To address this, our main insight is to optimize over a continuous superset of sparse matrices with a fixed structure known as products of butterfly matrices. As butterfly matrices are not hardware efficient, we propose simple variants of butterfly (block and flat) to take advantage of modern hardware. Our method (Pixelated Butterfly) uses a simple fixed sparsity pattern based on flat block butterfly and low-rank matrices to sparsify most network layers (e.g., attention, MLP). We empirically validate that Pixelated Butterfly is $3\times$ faster than Butterfly and speeds up training to achieve favorable accuracy--efficiency tradeoffs. On the ImageNet classification and WikiText-103 language modeling tasks, our sparse models train up to 2.3$\times$ faster than the dense MLP-Mixer, Vision Transformer, and GPT-2 small with no drop in accuracy. | accept |
Title: Modeling label correlations implicitly through latent label encodings for multi-label text classification. Abstract: Multi-label text classification (MLTC) aims to assign a set of labels to each given document. Unlike single-label text classification methods that often focus on document representation learning, MLTC faces a key challenge of modeling label correlations due to complex label dependencies. Previous state-of-the-art works model label correlations explicitly. It lacks flexibility and is prone to introduce inductive bias that may not always hold, such as label-correlation simplification, sequencing label sets, and label-correlation overload. To address this issue, this paper uses latent label representations to model label correlations implicitly. Specifically, the proposed method concatenates a set of latent labels (instead of actual labels) to the text tokens, inputs them to BERT, then maps the contextual encodings of these latent labels to actual labels cooperatively. The correlations between labels, and between labels and the text are modeled indirectly through these latent-label encodings and their correlations. Such latent and distributed correlation modeling can impose less a priori limits and provide more flexibility. The method is conceptually simple but quite effective. It improves the state-of-the-art results on two widely used benchmark datasets by a large margin. Further experiments demonstrate that its effectiveness lies in label-correlation utilization rather than document representation. Feature study reveals the importance of using latent label embeddings. It also reveals that contrary to the other token embeddings, the embeddings of these latent labels are sensitive to tasks; sometimes pretraining them can lead to significant performance loss rather than promotion. This result suggests that they are more related to task information (i.e., the actual labels) than the other tokens. | reject |
Title: Inference-Time Personalized Federated Learning. Abstract: In Federated learning (FL), multiple clients collaborate to learn a model through a central server but keep the data decentralized. Personalized federated learning (PFL) further extends FL to handle data heterogeneity between clients by learning personalized models. In both FL and PFL, all clients participate in the training process and their labeled data is used for training. However, in reality, novel clients may wish to join a prediction service after it has been deployed, obtaining predictions for their own unlabeled data.
Here, we defined a new learning setup, Inference-Time PFL (IT-PFL), where a model trained on a set of clients, needs to be later evaluated on novel unlabeled clients at inference time. We propose a novel approach to this problem IT-PFL-HN, based on a hypernetwork module and an encoder module. Specifically, we train an encoder network that learns a representation for a client given its unlabeled data. That client representation is fed to a hypernetwork that generates a personalized model for that client. Evaluated on four benchmark datasets, we find that IT-PFL-HN generalizes better than current FL and PFL methods, especially when the novel client has a large domain shift. We also analyzed the generalization error for the novel client, showing how it can be bounded using results from multi-task learning and domain adaptation. Finally, since novel clients do not contribute their data to training, they can potentially have better control over their data privacy; Indeed, we showed analytically and experimentally how novel clients can apply differential privacy to their data. | reject |
Title: Over-parameterization Improves Generalization in the XOR Detection Problem. Abstract: Empirical evidence suggests that neural networks with ReLU activations generalize better with over-parameterization. However, there is currently no theoretical analysis that explains this observation. In this work, we study a simplified learning task with over-parameterized convolutional networks that empirically exhibits the same qualitative phenomenon. For this setting, we provide a theoretical analysis of the optimization and generalization performance of gradient descent. Specifically, we prove data-dependent sample complexity bounds which show that over-parameterization improves the generalization performance of gradient descent. | reject |
Title: Distance-Based Learning from Errors for Confidence Calibration. Abstract: Deep neural networks (DNNs) are poorly calibrated when trained in conventional ways. To improve confidence calibration of DNNs, we propose a novel training method, distance-based learning from errors (DBLE). DBLE bases its confidence estimation on distances in the representation space. In DBLE, we first adapt prototypical learning to train classification models. It yields a representation space where the distance between a test sample and its ground truth class center can calibrate the model's classification performance. At inference, however, these distances are not available due to the lack of ground truth labels. To circumvent this by inferring the distance for every test sample, we propose to train a confidence model jointly with the classification model. We integrate this into training by merely learning from mis-classified training samples, which we show to be highly beneficial for effective learning. On multiple datasets and DNN architectures, we demonstrate that DBLE outperforms alternative single-model confidence calibration approaches. DBLE also achieves comparable performance with computationally-expensive ensemble approaches with lower computational cost and lower number of parameters. | accept |
Title: Beyond Classical Diffusion: Ballistic Graph Neural Network. Abstract: This paper presents the ballistic graph neural network. Ballistic graph neural network tackles the weight distribution from a transportation perspective and has many different properties comparing to the traditional graph neural network pipeline. The ballistic graph neural network does not require to calculate any eigenvalue. The filters propagate exponentially faster($\sigma^2 \sim T^2$) comparing to traditional graph neural network($\sigma^2 \sim T$). We use a perturbed coin operator to perturb and optimize the diffusion rate. Our results show that by selecting the diffusion speed, the network can reach a similar accuracy with fewer parameters. We also show the perturbed filters act as better representations comparing to pure ballistic ones. We provide a new perspective of training graph neural network, by adjusting the diffusion rate, the neural network's performance can be improved. | reject |
Title: DMSANET: DUAL MULTI SCALE ATTENTION NETWORK. Abstract: Attention mechanism of late has been quite popular in the computer vision community. A lot of work has been done to improve the performance of the network,
although almost always it results in increased computational complexity. In this paper, we propose a new attention module that not only achieves the best performance
but also has lesser parameters compared to most existing models. Our attention
module can easily be integrated with other convolutional neural networks because
of its lightweight nature. The proposed network named Dual Multi Scale Attention
Network (DMSANet) is comprised of two parts: the first part is used to extract
features at various scales and aggregate them, the second part uses spatial and
channel attention modules in parallel to adaptively integrate local features with
their global dependencies. We benchmark our network performance for Image
Classification on ImageNet dataset, Object Detection and Instance Segmentation
both on MS COCO dataset. | reject |
Title: Maximum Entropy Population Based Training for Zero-Shot Human-AI Coordination. Abstract: An AI agent should be able to coordinate with humans to solve tasks. We consider the problem of training a Reinforcement Learning (RL) agent without using any human data, i.e., in a zero-shot setting, to make it capable of collaborating with humans. Standard RL agents learn through self-play. Unfortunately, these agents only know how to collaborate with themselves and normally do not perform well with unseen partners, such as humans. The methodology of how to train a robust agent in a zero-shot fashion is still subject to research. Motivated from the maximum entropy RL, we derive a centralized population entropy objective to facilitate learning of a diverse population of agents, which is later used to train a robust AI agent to collaborate with unseen partners. The proposed method shows its effectiveness compared to baseline methods, including self-play PPO, the standard Population-Based Training (PBT), and trajectory diversity-based PBT, in the popular Overcooked game environment. We also conduct online experiments with real humans and further demonstrate the efficacy of the method in the real world. | reject |
Title: Robust Imitation via Mirror Descent Inverse Reinforcement Learning. Abstract: Adversarial imitation learning techniques are based on modeling statistical divergences using agent and expert demonstration data. However, unbiased minimization of these divergences is not usually guaranteed due to the geometry of the underlying space. Furthermore, when the size of demonstrations is not sufficient, estimated reward functions from the discriminative signals become uncertain and fail to give informative feedback. Instead of formulating a global cost at once, we consider reward functions as an iterative sequence in a proximal method. In this paper, we show that rewards dervied by mirror descent ensures minimization of a Bregman divergence in terms of a rigorous regret bound of $\mathcal{O}(1/T)$ for a particular condition of step sizes $\{\eta_t\}_{t=1}^T$. The resulting mirror descent adversarial inverse reinforcement learning (MD-AIRL) algorithm gradually advances a parameterized reward function in an associated reward space, and the sequence of such functions provides optimization targets for the policy space. We empirically validate our method in discrete and continuous benchmarks and show that MD-AIRL outperforms previous methods in various settings. | reject |
Title: Provable Adaptation across Multiway Domains via Representation Learning. Abstract: This paper studies zero-shot domain adaptation where each domain is indexed on a multi-dimensional array, and we only have data from a small subset of domains. Our goal is to produce predictors that perform well on \emph{unseen} domains. We propose a model which consists of a domain-invariant latent representation layer and a domain-specific linear prediction layer with a low-rank tensor structure. Theoretically, we present explicit sample complexity bounds to characterize the prediction error on unseen domains in terms of the number of domains with training data and the number of data per domain. To our knowledge, this is the first finite-sample guarantee for zero-shot domain adaptation. In addition, we provide experiments on two-way MNIST and four-way fiber sensing datasets to demonstrate the effectiveness of our proposed model. | accept |
Title: More Side Information, Better Pruning: Shared-Label Classification as a Case Study. Abstract: Pruning of neural networks, also known as compression or sparsification, is the task of converting a given network, which may be too expensive to use (in prediction) on low resource platforms, with another 'lean' network which performs almost as well as the original one, while using considerably fewer resources. By turning the compression ratio knob, the practitioner can trade off the information gain versus the necessary computational resources, where information gain is a measure of reduction of uncertainty in the prediction.
In certain cases, however, the practitioner may readily possess some information on the prediction from other sources. The main question we study here is, whether it is possible to take advantage of the additional side information, in order to further reduce the computational resources, in tandem with the pruning process?
Motivated by a real-world application, we distill the following elegantly stated problem. We are given a multi-class prediction problem, combined with a (possibly pre-trained) network architecture for solving it on a given instance distribution, and also a method for pruning the network to allow trading off prediction speed with accuracy. We assume the network and the pruning methods are state-of-the-art, and it is not our goal here to improve them. However, instead of being asked to predict a single drawn instance $x$, we are being asked to predict the label of an $n$-tuple of instances $(x_1,\dots x_n)$, with the additional side information of all tuple instances share the same label. The shared label distribution is identical to the distribution on which the network was trained.
One trivial way to do this is by obtaining individual raw predictions for each of the $n$ instances (separately), using our given network, pruned for a desired accuracy, then taking the average to obtain a single more accurate prediction. This is simple to implement but intuitively sub-optimal, because the $n$ independent instantiations of the network do not share any information, and would probably waste resources on overlapping computation.
We propose various methods for performing this task, and compare them using extensive experiments on public benchmark data sets for image classification. Our comparison is based on measures of relative information (RI) and $n$-accuracy, which we define. Interestingly, we empirically find that I) sharing information between the $n$ independently computed hidden representations of $x_1,..,x_n$, using an LSTM based gadget, performs best, among all methods we experiment with, ii) for all methods studied, we exhibit a sweet spot phenomenon, which sheds light on the compression-information trade-off and may assist a practitioner to choose the desired compression ratio. | reject |
Title: Multi-modal Self-Supervision from Generalized Data Transformations. Abstract: In the image domain, excellent representation can be learned by inducing invariance to content-preserving transformations, such as image distortions. In this paper, we show that, for videos, the answer is more complex, and that better results can be obtained by accounting for the interplay between invariance, distinctiveness, multiple modalities and time. We introduce Generalized Data Transformations (GDTs) as a way to capture this interplay. GDTs reduce most previous self-supervised approaches to a choice of data transformations, even when this was not the case in the original formulations. They also allow to choose whether the representation should be invariant or distinctive w.r.t. each effect and tell which combinations are valid, thus allowing us to explore the space of combinations systematically. We show in this manner that being invariant to certain transformations and distinctive to others is critical to learning effective video representations, improving the state-of-the-art by a large margin, and even surpassing supervised pretraining. We demonstrate results on a variety of downstream video and audio classification and retrieval tasks, on datasets such as HMDB-51, UCF-101, DCASE2014, ESC-50 and VGG-Sound. In particular, we achieve new state-of-the-art accuracies of 72.8% on HMDB-51 and 95.2% on UCF-101. | reject |
Title: Stackelberg GAN: Towards Provable Minimax Equilibrium via Multi-Generator Architectures. Abstract: We study the problem of alleviating the instability issue in the GAN training procedure via new architecture design. The discrepancy between the minimax and maximin objective values could serve as a proxy for the difficulties that the alternating gradient descent encounters in the optimization of GANs. In this work, we give new results on the benefits of multi-generator architecture of GANs. We show that the minimax gap shrinks to \epsilon as the number of generators increases with rate O(1/\epsilon). This improves over the best-known result of O(1/\epsilon^2). At the core of our techniques is a novel application of Shapley-Folkman lemma to the generic minimax problem, where in the literature the technique was only known to work when the objective function is restricted to the Lagrangian function of a constraint optimization problem. Our proposed Stackelberg GAN performs well experimentally in both synthetic and real-world datasets, improving Frechet Inception Distance by 14.61% over the previous multi-generator GANs on the benchmark datasets. | reject |
Title: Fantastic Generalization Measures and Where to Find Them. Abstract: Generalization of deep networks has been intensely researched in recent years, resulting in a number of theoretical bounds and empirically motivated measures. However, most papers proposing such measures only study a small set of models, leaving open the question of whether these measures are truly useful in practice. We present the first large scale study of generalization bounds and measures in deep networks. We train over two thousand CIFAR-10 networks with systematic changes in important hyper-parameters. We attempt to uncover potential causal relationships between each measure and generalization, by using rank correlation coefficient and its modified forms. We analyze the results and show that some of the studied measures are very promising for further research. | accept |
Title: Understanding Generalization in Recurrent Neural Networks. Abstract: In this work, we develop the theory for analyzing the generalization performance of recurrent neural networks. We first present a new generalization bound for recurrent neural networks based on matrix 1-norm and Fisher-Rao norm. The definition of Fisher-Rao norm relies on a structural lemma about the gradient of RNNs. This new generalization bound assumes that the covariance matrix of the input data is positive definite, which might limit its use in practice. To address this issue, we propose to add random noise to the input data and prove a generalization bound for training with random noise, which is an extension of the former one. Compared with existing results, our generalization bounds have no explicit dependency on the size of networks. We also discover that Fisher-Rao norm for RNNs can be interpreted as a measure of gradient, and incorporating this gradient measure not only can tighten the bound, but allows us to build a relationship between generalization and trainability. Based on the bound, we theoretically analyze the effect of covariance of features on generalization of RNNs and discuss how weight decay and gradient clipping in the training can help improve generalization. | accept |
Title: Integer Networks for Data Compression with Latent-Variable Models. Abstract: We consider the problem of using variational latent-variable models for data compression. For such models to produce a compressed binary sequence, which is the universal data representation in a digital world, the latent representation needs to be subjected to entropy coding. Range coding as an entropy coding technique is optimal, but it can fail catastrophically if the computation of the prior differs even slightly between the sending and the receiving side. Unfortunately, this is a common scenario when floating point math is used and the sender and receiver operate on different hardware or software platforms, as numerical round-off is often platform dependent. We propose using integer networks as a universal solution to this problem, and demonstrate that they enable reliable cross-platform encoding and decoding of images using variational models. | accept |
Title: Ensembles of Generative Adversarial Networks for Disconnected Data. Abstract: Most computer vision datasets are composed of disconnected sets, such as images of different objects. We prove that distributions of this type of data cannot be represented with a continuous generative network without error, independent of the learning algorithm used. Disconnected datasets can be represented in two ways: with an ensemble of networks or with a single network using a truncated latent space. We show that ensembles are more desirable than truncated distributions for several theoretical and computational reasons. We construct a regularized optimization problem that rigorously establishes the relationships between a single continuous GAN, an ensemble of GANs, conditional GANs, and Gaussian Mixture GANs. The regularization can be computed efficiently, and we show empirically that our framework has a performance sweet spot that can be found via hyperparameter tuning. The ensemble framework provides better performance than a single continuous GAN or cGAN while maintaining fewer total parameters. | reject |
Title: SERCNN: Stacked Embedding Recurrent Convolutional Neural Network in Depression Detection on Twitter. Abstract: Conventional approach of self-reporting-based screening for depression is not scalable, expensive, and requires one to be fully aware of their mental health. Motivated by previous studies that demonstrated great potentials for using social media posts to monitor and predict one's mental health status, this study utilizes natural language processing and machine learning techniques on social media data to predict one's risk of depression. Most existing works utilize handcrafted features, and the adoption of deep learning in this domain is still lacking. Social media texts are often unstructured, ill-formed, and contain typos, making handcrafted features and conventional feature extraction methods inefficient. Moreover, prediction models built on these features often require a high number of posts per individual for accurate predictions. Therefore, this study proposes a Stacked Embedding Recurrent Convolutional Neural Network (SERCNN) for a more optimized prediction that has a better trade-off between the number of posts and accuracy. Feature vectors of two widely available pretrained embeddings trained on two distinct datasets are stacked, forming a meta-embedding vector that has a more robust and richer representation for any given word. We adapt Lai et al. (2015) RCNN approach that incorporates both the embedding vector and context learned from the neural network to form the final user representation before performing classification. We conducted our experiments on the Shen et al. (2017) depression Twitter dataset, the largest ground truth dataset used in this domain. Using SERCNN, our proposed model achieved a prediction accuracy of 78% when using only ten posts from each user, and the accuracy increases to 90% with an F1-measure of 0.89 when five hundred posts are analyzed. | reject |
Title: CDT: Cascading Decision Trees for Explainable Reinforcement Learning. Abstract: Deep Reinforcement Learning (DRL) has recently achieved significant advances in various domains. However, explaining the policy of RL agents still remains an open problem due to several factors, one being the complexity of explaining neural networks decisions. Recently, a group of works have used decision-tree-based models to learn explainable policies. Soft decision trees (SDTs) and discretized differentiable decision trees (DDTs) have been demonstrated to achieve both good performance and share the benefit of having explainable policies. In this work, we further improve the results for tree-based explainable RL in both performance and explainability. Our proposal, Cascading Decision Trees (CDTs) apply representation learning on the decision path to allow richer expressivity. Empirical results show that in both situations, where CDTs are used as policy function approximators or as imitation learners to explain black-box policies, CDTs can achieve better performances with more succinct and explainable models than SDTs. As a second contribution our study reveals limitations of explaining black-box policies via imitation learning with tree-based explainable models, due to its inherent instability. | reject |
Title: Information Bottleneck: Exact Analysis of (Quantized) Neural Networks. Abstract: The information bottleneck (IB) principle has been suggested as a way to analyze deep neural networks. The learning dynamics are studied by inspecting the mutual information (MI) between the hidden layers and the input and output. Notably, separate fitting and compression phases during training have been reported. This led to some controversy including claims that the observations are not reproducible and strongly dependent on the type of activation function used as well as on the way the MI is estimated. Our study confirms that different ways of binning when computing the MI lead to qualitatively different results, either supporting or refusing IB conjectures.
To resolve the controversy, we study the IB principle in settings where MI is non-trivial and can be computed exactly. We monitor the dynamics of quantized neural networks, that is, we discretize the whole deep learning system so that no approximation is required when computing the MI. This allows us to quantify the information flow without measurement errors.
In this setting, we observed a fitting phase for all layers and a compression phase for the output layer in all experiments; the compression in the hidden layers was dependent on the type of activation function. Our study shows that the initial IB results were not artifacts of binning when computing the MI. However, the critical claim that the compression phase may not be observed for some networks also holds true. | accept |
Title: Towards More Realistic Neural Network Uncertainties. Abstract: Statistical models are inherently uncertain. Quantifying or at least upper-bounding their uncertainties is vital for safety-critical systems. While standard neural networks do not report this information, several approaches exist to integrate uncertainty estimates into them. Assessing the quality of these uncertainty estimates is not straightforward, as no direct ground truth labels are available. Instead, implicit statistical assessments are required. For regression, we propose to evaluate uncertainty realism---a strict quality criterion---with a Mahalanobis distance-based statistical test. An empirical evaluation reveals the need for uncertainty measures that are appropriate to upper-bound heavy-tailed empirical errors. Alongside, we transfer the variational U-Net classification architecture to standard supervised image-to-image tasks. It provides two uncertainty mechanisms and significantly improves uncertainty realism compared to a plain encoder-decoder model. | reject |
Title: Momentum Conserving Lagrangian Neural Networks. Abstract: Realistic models of physical world rely on differentiable symmetries that, in turn, correspond to conservation laws. Recent works on Lagrangian and Hamiltonian neural networks show that the underlying symmetries of a system can be easily learned by a neural network when provided with an appropriate inductive bias. However, these models still suffer from issues such as inability to generalize to arbitrary system sizes, poor interpretability, and most importantly, inability to learn translational and rotational symmetries, which lead to the conservation laws of linear and angular momentum, respectively. Here, we present a momentum conserving Lagrangian neural network (MCLNN) that learns the Lagrangian of a system, while also preserving the translational and rotational symmetries. We test our approach on linear and non-linear spring systems, and a gravitational system, demonstrating the energy and momentum conservation. We also show that the model developed can generalize to systems of any arbitrary size. Finally, we discuss the interpretability of the MCLNN, which directly provides physical insights into the interactions of multi-particle systems. | reject |
Title: A Universal Music Translation Network. Abstract: We present a method for translating music across musical instruments and styles. This method is based on unsupervised training of a multi-domain wavenet autoencoder, with a shared encoder and a domain-independent latent space that is trained end-to-end on waveforms. Employing a diverse training dataset and large net capacity, the single encoder allows us to translate also from musical domains that were not seen during training. We evaluate our method on a dataset collected from professional musicians, and achieve convincing translations. We also study the properties of the obtained translation and demonstrate translating even from a whistle, potentially enabling the creation of instrumental music by untrained humans. | accept |
Title: Label Refining: a semi-supervised method to extract voice characteristics without ground truth. Abstract: A characteristic is a distinctive trait shared by a group of observations which may be used to identify them. In the context of voice casting for audiovisual productions, characteristic extraction has an important role since it can help explaining the decisions of a voice recommendation system, or give modalities to the user with the aim to express a voice search request. Unfortunately, the lack of standard taxonomy to describe comedian voices prevents the implementation of an annotation protocol. To address this problem, we propose a new semi-supervised learning method entitled Label Refining that consists in extracting refined labels (e.g. vocal characteristics) from known initial labels (e.g. character played in a recording). Our proposed method first suggests using a representation extractor based on the initial labels, then computing refined labels using a clustering algorithm to finally train a refined representation extractor. The method is validated by applying Label Refining on recordings from the video game MassEffect 3. Experiments show that, using a subsidiary corpus, it is possible to bring out interesting voice characteristics without any a priori knowledge. | reject |
Title: Learning continuous-time PDEs from sparse data with graph neural networks. Abstract: The behavior of many dynamical systems follow complex, yet still unknown partial differential equations (PDEs). While several machine learning methods have been proposed to learn PDEs directly from data, previous methods are limited to discrete-time approximations or make the limiting assumption of the observations arriving at regular grids. We propose a general continuous-time differential model for dynamical systems whose governing equations are parameterized by message passing graph neural networks. The model admits arbitrary space and time discretizations, which removes constraints on the locations of observation points and time intervals between the observations. The model is trained with continuous-time adjoint method enabling efficient neural PDE inference. We demonstrate the model's ability to work with unstructured grids, arbitrary time steps, and noisy observations. We compare our method with existing approaches on several well-known physical systems that involve first and higher-order PDEs with state-of-the-art predictive performance. | accept |
Title: On the Convergence of Projected Alternating Maximization for Equitable and Optimal Transport. Abstract: This paper studies the equitable and optimal transport (EOT) problem, which has many applications such as fair division problems and optimal transport with multiple agents etc. In the discrete distributions case, the EOT problem can be formulated as a linear program (LP). Since this LP is prohibitively large for general LP solvers, Scetbon \etal suggests to perturb the problem by adding an entropy regularization. They proposed a projected alternating maximization algorithm (PAM) to solve the dual of the entropy regularized EOT. In this paper, we provide the first convergence analysis of PAM. A novel rounding procedure is proposed to help construct the primal solution for the original EOT problem. We also propose a variant of PAM by incorporating the extrapolation technique that can numerically improve the performance of PAM. Results in this paper may shed lights on block coordinate (gradient) descent methods for general optimization problems. | reject |
Title: Are wider nets better given the same number of parameters?. Abstract: Empirical studies demonstrate that the performance of neural networks improves with increasing number of parameters. In most of these studies, the number of parameters is increased by increasing the network width. This begs the question: Is the observed improvement due to the larger number of parameters, or is it due to the larger width itself? We compare different ways of increasing model width while keeping the number of parameters constant. We show that for models initialized with a random, static sparsity pattern in the weight tensors, network width is the determining factor for good performance, while the number of weights is secondary, as long as the model achieves high training accuarcy. As a step towards understanding this effect, we analyze these models in the framework of Gaussian Process kernels. We find that the distance between the sparse finite-width model kernel and the infinite-width kernel at initialization is indicative of model performance. | accept |
Title: Sparse Communication via Mixed Distributions. Abstract: Neural networks and other machine learning models compute continuous representations, while humans communicate mostly through discrete symbols. Reconciling these two forms of communication is desirable for generating human-readable interpretations or learning discrete latent variable models, while maintaining end-to-end differentiability. Some existing approaches (such as the Gumbel-Softmax transformation) build continuous relaxations that are discrete approximations in the zero-temperature limit, while others (such as sparsemax transformations and the Hard Concrete distribution) produce discrete/continuous hybrids. In this paper, we build rigorous theoretical foundations for these hybrids, which we call "mixed random variables.'' Our starting point is a new "direct sum'' base measure defined on the face lattice of the probability simplex. From this measure, we introduce new entropy and Kullback-Leibler divergence functions that subsume the discrete and differential cases and have interpretations in terms of code optimality. Our framework suggests two strategies for representing and sampling mixed random variables, an extrinsic ("sample-and-project'’) and an intrinsic one (based on face stratification). We experiment with both approaches on an emergent communication benchmark and on modeling MNIST and Fashion-MNIST data with variational auto-encoders with mixed latent variables. | accept |
Title: Modelling neuronal behaviour with time series regression: Recurrent Neural Networks on synthetic C. elegans data. Abstract: Given the inner complexity of the human nervous system, insight into the dynamics of brain activity can be gained from understanding smaller and simpler organisms, such as the nematode C. elegans. The behavioural and structural biology of these organisms is well-known, making them prime candidates for benchmarking modelling and simulation techniques. In these complex neuronal collections, classical white-box modelling techniques based on intrinsic structural or behavioural information are either unable to capture the profound nonlinearities of the neuronal response to different stimuli or generate extremely complex models, which are computationally intractable. In this paper we investigate whether it is possible to generate lower complexity black-box models that can capture the system dynamics with low error using only measured or simulated input-output information. We show how the nervous system of C. elegans can be modelled and simulated with data-driven models using different neural network architectures. Specifically, we target the use of state of the art recurrent neural networks architectures such as LSTMs and GRUs and compare these architectures in terms of their properties and their RMSE, as well as the complexity of the resulting models.
We show that GRU models with a hidden layer size of 4 units are able to accurately reproduce the system's response to very different stimuli. | reject |
Title: Graph Deformer Network. Abstract: Convolution learning on graphs draws increasing attention recently due to its potential applications to a large amount of irregular data. Most graph convolution methods leverage the plain summation/average aggregation to avoid the discrepancy of responses from isomorphic graphs. However, such an extreme collapsing way would result in a structural loss and signal entanglement of nodes, which further cause the degradation of the learning ability. In this paper, we propose a simple yet effective graph deformer network (GDN) to fulfill anisotropic convolution filtering on graphs, analogous to the standard convolution operation on images. Local neighborhood subgraphs (acting like receptive fields) with different structures are deformed into a unified virtual space, coordinated by several anchor nodes. In space deformation, we transfer components of nodes therein into anchors by learning their correlation, and build a pseudo multi-granularity plane calibrated with anchors. Anisotropic convolutional kernels can be further performed over the anchor-coordinated space to well encode local variations of receptive fields. By parameterizing anchors and stacking coarsening layers, we build a graph deformer network in an end-to-end fashion. Theoretical analysis indicates its connection to previous work and shows the promising property of isomorphism testing. Extensive experiments on widely-used datasets validate the effectiveness of the proposed GDN in node and graph classifications. | reject |
Title: Self-supervised regression learning using domain knowledge: Applications to improving self-supervised image denoising. Abstract: Regression that predicts continuous quantity is a central part of applications using computational imaging and computer vision technologies. Yet, studying and understanding self-supervised learning for regression tasks -- except for a particular regression task, image denoising -- have lagged behind. This paper proposes a general self-supervised regression learning (SSRL) framework that enables learning regression neural networks with only input data (but without ground-truth target data), by using a designable operator that encapsulates domain knowledge of a specific application. The paper underlines the importance of domain knowledge by showing that under some mild conditions, the better designable operator is used, the proposed SSRL loss becomes closer to ordinary supervised learning loss. Numerical experiments for natural image denoising and low-dose computational tomography denoising demonstrate that proposed SSRL significantly improves the denoising quality over several existing self-supervised denoising methods. | reject |
Title: KL Guided Domain Adaptation. Abstract: Domain adaptation is an important problem and often needed for real-world applications. In this problem, instead of i.i.d. training and testing datapoints, we assume that the source (training) data and the target (testing) data have different distributions. With that setting, the empirical risk minimization training procedure often does not perform well, since it does not account for the change in the distribution. A common approach in the domain adaptation literature is to learn a representation of the input that has the same (marginal) distribution over the source and the target domain. However, these approaches often require additional networks and/or optimizing an adversarial (minimax) objective, which can be very expensive or unstable in practice. To improve upon these marginal alignment techniques, in this paper, we first derive a generalization bound for the target loss based on the training loss and the reverse Kullback-Leibler (KL) divergence between the source and the target representation distributions. Based on this bound, we derive an algorithm that minimizes the KL term to obtain a better generalization to the target domain. We show that with a probabilistic representation network, the KL term can be estimated efficiently via minibatch samples without any additional network or a minimax objective. This leads to a theoretically sound alignment method which is also very efficient and stable in practice. Experimental results also suggest that our method outperforms other representation-alignment approaches. | accept |
Title: Molecular Graph Representation Learning via Heterogeneous Motif Graph Construction. Abstract: We consider feature representation learning of molecular graphs. Graph Neural Networks have been widely used in feature representation learning of molecular graphs. However, most proposed methods focus on the individual molecular graph while neglecting their connections, such as motif-level relationships. We propose a novel molecular graph representation learning method by constructing a Heterogeneous Motif graph (HM-graph) to address this issue. In particular, we build an HM-graph that contains motif nodes and molecular nodes. Each motif node corresponds to a motif extracted from molecules. Then, we propose a Heterogeneous Motif Graph Neural Network (HM-GNN) to learn feature representations for each node in the HM-graph. Our HM-graph also enables effective multi-task learning, especially for small molecular datasets. To address the potential efficiency issue, we propose an edge sampler, which significantly reduces computational resources usage. The experimental results show that our model consistently outperforms previous state-of-the-art models. Under multi-task settings, the promising performances of our methods on combined datasets shed light on a new learning paradigm for small molecular datasets. Finally, we show that our model achieves similar performances with significantly less computational resources by using our edge sampler. | reject |
Title: Conditional Expectation based Value Decomposition for Scalable On-Demand Ride Pooling. Abstract: Owing to the benefits for customers (lower prices), drivers (higher revenues), aggregation companies (higher revenues) and the environment (fewer vehicles), on-demand ride pooling (e.g., Uber pool, Grab Share) has become quite popular. The significant computational complexity of matching vehicles to combinations of requests has meant that traditional ride pooling approaches are myopic in that they do not consider the impact of current matches on future value for vehicles/drivers.
Recently, Neural Approximate Dynamic Programming (NeurADP) has employed value decomposition with Approximate Dynamic Programming (ADP) to outperform leading approaches by considering the impact of an individual agent's (vehicle) chosen actions on the future value of that agent. However, in order to ensure scalability and facilitate city-scale ride pooling, NeurADP completely ignores the impact of other agents actions on individual agent/vehicle value. As demonstrated in our experimental results, ignoring the impact of other agents actions on individual value can have a significant impact on the overall performance when there is increased competition among vehicles for demand. Our key contribution is a novel mechanism based on computing conditional expectations through joint conditional probabilities for capturing dependencies on other agents actions without increasing the complexity of training or decision making. We show that our new approach, Conditional Expectation based Value Decomposition (CEVD) outperforms NeurADP by up to 9.76$\% $in terms of overall requests served, which is a significant improvement on a city wide benchmark taxi dataset. | reject |
Title: DASGrad: Double Adaptive Stochastic Gradient. Abstract: Adaptive moment methods have been remarkably successful for optimization under the presence of high dimensional or sparse gradients, in parallel to this, adaptive sampling probabilities for SGD have allowed optimizers to improve convergence rates by prioritizing examples to learn efficiently. Numerous applications in the past have implicitly combined adaptive moment methods with adaptive probabilities yet the theoretical guarantees of such procedures have not been explored. We formalize double adaptive stochastic gradient methods DASGrad as an optimization technique and analyze its convergence improvements in a stochastic convex optimization setting, we provide empirical validation of our findings with convex and non convex objectives. We observe that the benefits of the method increase with the model complexity and variability of the gradients, and we explore the resulting utility in extensions to transfer learning. | reject |
Title: Offline Reinforcement Learning with Resource Constrained Online Deployment. Abstract: Offline reinforcement learning is used to train policies in scenarios where real-time access to the environment is expensive or impossible.
As a natural consequence of these harsh conditions, an agent may lack the resources to fully observe the online environment before taking an action. We dub this situation the resource-constrained setting. This leads to situations where the offline dataset (available for training) can contain fully processed features (using powerful language models, image models, complex sensors, etc.) which are not available when actions are actually taken online.
This disconnect leads to an interesting and unexplored problem in offline RL: Is it possible to use a richly processed offline dataset to train a policy which has access to fewer features in the online environment?
In this work, we introduce and formalize this novel resource-constrained problem setting. We highlight the performance gap between policies trained using the full offline dataset and policies trained using limited features.
We address this performance gap with a policy transfer algorithm which first trains a teacher agent using the offline dataset where features are fully available, and then transfers this knowledge to a student agent that only uses the resource-constrained features. To better capture the challenge of this setting, we propose a data collection procedure: Resource Constrained-Datasets for RL (RC-D4RL). We evaluate our transfer algorithm on RC-D4RL and the popular D4RL benchmarks and observe consistent improvement over the baseline (TD3+BC without transfer). | reject |
Title: Second-Moment Loss: A Novel Regression Objective for Improved Uncertainties. Abstract: Quantification of uncertainty is one of the most promising approaches to establish safe machine learning. Despite its importance, it is far from being generally solved, especially for neural networks. One of the most commonly used approaches so far is Monte Carlo dropout, which is computationally cheap and easy to apply in practice. However, it can underestimate the uncertainty. We propose a new objective, referred to as second-moment loss (SML), to address this issue. While the full network is encouraged to model the mean, the dropout networks are explicitly used to optimize the model variance. We analyze the performance of the new objective on various toy and UCI regression datasets. Comparing to the state-of-the-art of deep ensembles, SML leads to comparable prediction accuracies and uncertainty estimates while only requiring a single model. Under distribution shift, we observe moderate improvements. From a safety perspective also the study of worst-case uncertainties is crucial. In this regard we improve considerably. Finally, we show that SML can be successfully applied to SqueezeDet, a modern object detection network. We improve on its uncertainty-related scores while not deteriorating regression quality. As a side result, we introduce an intuitive Wasserstein distance-based uncertainty measure that is non-saturating and thus allows to resolve quality differences between any two uncertainty estimates. | reject |
Title: Subspace Regularizers for Few-Shot Class Incremental Learning. Abstract: Few-shot class incremental learning---the problem of updating a trained classifier to discriminate among an expanded set of classes with limited labeled data---is a key challenge for machine learning systems deployed in non-stationary environments. Existing approaches to the problem rely on complex model architectures and training procedures that are difficult to tune and re-use. In this paper, we present an extremely simple approach that enables the use of ordinary logistic regression classifiers for few-shot incremental learning. The key to this approach is a new family of \textit{subspace regularization} schemes that encourage weight vectors for new classes to lie close to the subspace spanned by the weights of existing classes. When combined with pretrained convolutional feature extractors, logistic regression models trained with subspace regularization outperform specialized, state-of-the-art approaches to few-shot incremental image classification by up to 23\% on the \textit{mini}ImageNet dataset. Because of its simplicity, subspace regularization can be straightforwardly configured to incorporate additional background information about the new classes (including class names and descriptions specified in natural language); this offers additional control over the trade-off between existing and new classes. Our results show that simple geometric regularization of class representations offers an effective tool for continual learning. | accept |
Title: Understanding and Mitigating Accuracy Disparity in Regression. Abstract: With the widespread deployment of large-scale prediction systems in high-stakes domains, e.g., face recognition, criminal justice, etc., disparity on prediction accuracy between different demographic subgroups has called for fundamental understanding on the source of such disparity and algorithmic intervention to mitigate it. In this paper, we study the accuracy disparity problem in regression. To begin with, we first propose an error decomposition theorem, which decomposes the accuracy disparity into the distance between label populations and the distance between conditional representations, to help explain why such accuracy disparity appears in practice. Motivated by this error decomposition and the general idea of distribution alignment with statistical distances, we then propose an algorithm to reduce this disparity, and analyze its game-theoretic optima of the proposed objective function. We conduct experiments on four real-world datasets. The experimental results suggest that our proposed algorithms can effectively mitigate accuracy disparity while maintaining the predictive power of the regression models. | reject |
Title: Cost-Sensitive Robustness against Adversarial Examples. Abstract: Several recent works have developed methods for training classifiers that are certifiably robust against norm-bounded adversarial perturbations. These methods assume that all the adversarial transformations are equally important, which is seldom the case in real-world applications. We advocate for cost-sensitive robustness as the criteria for measuring the classifier's performance for tasks where some adversarial transformation are more important than others. We encode the potential harm of each adversarial transformation in a cost matrix, and propose a general objective function to adapt the robust training method of Wong & Kolter (2018) to optimize for cost-sensitive robustness. Our experiments on simple MNIST and CIFAR10 models with a variety of cost matrices show that the proposed approach can produce models with substantially reduced cost-sensitive robust error, while maintaining classification accuracy. | accept |
Title: Eidetic 3D LSTM: A Model for Video Prediction and Beyond. Abstract: Spatiotemporal predictive learning, though long considered to be a promising self-supervised feature learning method, seldom shows its effectiveness beyond future video prediction. The reason is that it is difficult to learn good representations for both short-term frame dependency and long-term high-level relations. We present a new model, Eidetic 3D LSTM (E3D-LSTM), that integrates 3D convolutions into RNNs. The encapsulated 3D-Conv makes local perceptrons of RNNs motion-aware and enables the memory cell to store better short-term features. For long-term relations, we make the present memory state interact with its historical records via a gate-controlled self-attention module. We describe this memory transition mechanism eidetic as it is able to effectively recall the stored memories across multiple time stamps even after long periods of disturbance. We first evaluate the E3D-LSTM network on widely-used future video prediction datasets and achieve the state-of-the-art performance. Then we show that the E3D-LSTM network also performs well on the early activity recognition to infer what is happening or what will happen after observing only limited frames of video. This task aligns well with video prediction in modeling action intentions and tendency. | accept |
Title: Tight Second-Order Certificates for Randomized Smoothing. Abstract: Randomized smoothing is a popular way of providing robustness guarantees against adversarial attacks: randomly-smoothed functions have a universal Lipschitz-like bound, allowing for robustness certificates to be easily computed. In this work, we show that there also exists a universal curvature-like bound for Gaussian random smoothing: given the exact value and gradient of a smoothed function, we compute a lower bound on the distance of a point to its closest adversarial example, called the Second-order Smoothing (SoS) robustness certificate. In addition to proving the correctness of this novel certificate, we show that SoS certificates are realizable and therefore tight. Interestingly, we show that the maximum achievable benefits, in terms of certified robustness, from using the additional information of the gradient norm are relatively small: because our bounds are tight, this is a fundamental negative result. The gain of SoS certificates further diminishes if we consider the estimation error of the gradient norms, for which we have developed an estimator. We therefore additionally develop a variant of Gaussian smoothing, called Gaussian dipole smoothing, which provides similar bounds to randomized smoothing with gradient information, but with much-improved sample efficiency. This allows us to achieve (marginally) improved robustness certificates on high-dimensional datasets such as CIFAR-10 and ImageNet. | reject |
Title: Permutation Compressors for Provably Faster Distributed Nonconvex Optimization. Abstract: In this work we study the MARINA method of Gorbunov et al (ICML, 2021) -- the current state-of-the-art distributed non-convex optimization method in terms of theoretical communication complexity. Theoretical superiority of this method can be largely attributed to two sources: a carefully engineered biased stochastic gradient estimator, which leads to a reduction in the number of communication rounds, and the reliance on
{\em independent} stochastic communication compression, which leads to a reduction in the number of transmitted bits within each communication round. In this paper we i) extend the theory of MARINA to support a much wider class of potentially {\em correlated} compressors, extending the reach of the method beyond the classical independent compressors setting, ii) show that a new quantity, for which we coin the name {\em Hessian variance}, allows us to significantly refine the original analysis of MARINA without any additional assumptions, and iii) identify a special class of correlated compressors based on the idea of {\em random permutations}, for which we coin the term Perm$K$, the use of which leads to up to $O(\sqrt{n})$ (resp. $O(1 + d/\sqrt{n})$) improvement in the theoretical communication complexity of MARINA in the low Hessian variance regime when $d\geq n$ (resp. $d \leq n$), where $n$ is the number of workers and $d$ is the number of parameters describing the model we are learning. We corroborate our theoretical results with carefully engineered synthetic experiments with minimizing the average of nonconvex quadratics, and on autoencoder training with the MNIST dataset. | accept |
Title: TOWARDS NATURAL ROBUSTNESS AGAINST ADVERSARIAL EXAMPLES. Abstract: Recent studies have shown that deep neural networks are vulnerable to adversarial examples, but most of the methods proposed to defense adversarial examples cannot solve this problem fundamentally. In this paper, we theoretically prove that there is an upper bound for neural networks with identity mappings to constrain the error caused by adversarial noises. However, in actual computations, this kind of neural network no longer holds any upper bound and is therefore susceptible to adversarial examples. Following similar procedures, we explain why adversarial examples can fool other deep neural networks with skip connections. Furthermore, we demonstrate that a new family of deep neural networks called Neural ODEs (Chen et al., 2018) holds a weaker upper bound. This weaker upper bound prevents the amount of change in the result from being too large. Thus, Neural ODEs have natural robustness against adversarial examples. We evaluate the performance of Neural ODEs compared with ResNet under three white-box adversarial attacks (FGSM, PGD, DI2-FGSM) and one black-box adversarial attack (Boundary Attack). Finally, we show that the natural robustness of Neural ODEs is even better than the robustness of neural networks that are trained with adversarial training methods, such as TRADES and YOPO.
| reject |
Title: Empowering Graph Representation Learning with Paired Training and Graph Co-Attention. Abstract: Through many recent advances in graph representation learning, performance achieved on tasks involving graph-structured data has substantially increased in recent years---mostly on tasks involving node-level predictions. The setup of prediction tasks over entire graphs (such as property prediction for a molecule, or side-effect prediction for a drug), however, proves to be more challenging, as the algorithm must combine evidence about several structurally relevant patches of the graph into a single prediction.
Most prior work attempts to predict these graph-level properties while considering only one graph at a time---not allowing the learner to directly leverage structural similarities and motifs across graphs. Here we propose a setup in which a graph neural network receives pairs of graphs at once, and extend it with a co-attentional layer that allows node representations to easily exchange structural information across them. We first show that such a setup provides natural benefits on a pairwise graph classification task (drug-drug interaction prediction), and then expand to a more generic graph regression setup: enhancing predictions over QM9, a standard molecular prediction benchmark. Our setup is flexible, powerful and makes no assumptions about the underlying dataset properties, beyond anticipating the existence of multiple training graphs. | reject |
Title: Unsupervised Meta-Learning through Latent-Space Interpolation in Generative Models. Abstract: Several recently proposed unsupervised meta-learning approaches rely on synthetic meta-tasks created using techniques such as random selection, clustering and/or augmentation. In this work, we describe a novel approach that generates meta-tasks using generative models. The proposed family of algorithms generate pairs of in-class and out-of-class samples from the latent space in a principled way, allowing us to create synthetic classes forming the training and validation data of a meta-task. We find that the proposed approach, LAtent Space Interpolation Unsupervised Meta-learning (LASIUM), outperforms or is competitive with current unsupervised learning baselines on few-shot classification tasks on the most widely used benchmark datasets. | accept |
Title: Learning to Generate Questions by Recovering Answer-containing Sentences. Abstract: To train a question answering model based on machine reading comprehension (MRC), significant effort is required to prepare annotated training data composed of questions and their answers from contexts. To mitigate this issue, recent research has focused on synthetically generating a question from a given context and an annotated (or generated) answer by training an additional generative model, which can be utilized to augment the training data. In light of this research direction, we propose a novel pre-training approach that learns to generate contextually rich questions, by recovering answer-containing sentences. Our approach is composed of two novel components, (1) dynamically determining K answers from a given document and (2) pre-training the question generator on the task of generating the answer-containing sentence. We evaluate our method against existing ones in terms of the quality of generated questions as well as the fine-tuned MRC model accuracy after training on the data synthetically generated by our method. Experimental results demonstrate that our approach consistently improves the question generation capability of existing models such as T5 and UniLM, and shows state-of-the-art results on MS MARCO and NewsQA, and comparable results to the state-of-the-art on SQuAD. Additionally, we demonstrate that the data synthetically generated by our approach is beneficial for boosting up the downstream MRC accuracy across a wide range of datasets, such as SQuAD-v1.1, v2.0, and KorQuAD, without any modification to the existing MRC models. Furthermore, our experiments highlight that our method shines especially when a limited amount of training data is given, in terms of both pre-training and downstream MRC data. | reject |
Title: Learning to cluster in order to transfer across domains and tasks. Abstract: This paper introduces a novel method to perform transfer learning across domains and tasks, formulating it as a problem of learning to cluster. The key insight is that, in addition to features, we can transfer similarity information and this is sufficient to learn a similarity function and clustering network to perform both domain adaptation and cross-task transfer learning. We begin by reducing categorical information to pairwise constraints, which only considers whether two instances belong to the same class or not (pairwise semantic similarity). This similarity is category-agnostic and can be learned from data in the source domain using a similarity network. We then present two novel approaches for performing transfer learning using this similarity function. First, for unsupervised domain adaptation, we design a new loss function to regularize classification with a constrained clustering loss, hence learning a clustering network with the transferred similarity metric generating the training inputs. Second, for cross-task learning (i.e., unsupervised clustering with unseen categories), we propose a framework to reconstruct and estimate the number of semantic clusters, again using the clustering network. Since the similarity network is noisy, the key is to use a robust clustering algorithm, and we show that our formulation is more robust than the alternative constrained and unconstrained clustering approaches. Using this method, we first show state of the art results for the challenging cross-task problem, applied on Omniglot and ImageNet. Our results show that we can reconstruct semantic clusters with high accuracy. We then evaluate the performance of cross-domain transfer using images from the Office-31 and SVHN-MNIST tasks and present top accuracy on both datasets. Our approach doesn't explicitly deal with domain discrepancy. If we combine with a domain adaptation loss, it shows further improvement. | accept |
Title: Sample and Communication-Efficient Decentralized Actor-Critic Algorithms with Finite-Time Analysis. Abstract: Actor-critic (AC) algorithms have been widely adopted in decentralized multi-agent systems to learn the optimal joint control policy. However, existing decentralized AC algorithms either do not preserve the privacy of agents or are not sample and communication-efficient. In this work, we develop two decentralized AC and natural AC (NAC) algorithms that are private, and sample and communication-efficient. In both algorithms, agents share noisy information to preserve privacy and adopt mini-batch updates to improve sample and communication efficiency. Particularly for decentralized NAC, we develop a decentralized Markovian SGD algorithm with an adaptive mini-batch size to efficiently compute the natural policy gradient. Under Markovian sampling and linear function approximation, we prove the proposed decentralized AC and NAC algorithms achieve the state-of-the-art sample complexities $\mathcal{O}(\epsilon^{-2}\ln\epsilon^{-1})$ and $\mathcal{O}(\epsilon^{-3}\ln\epsilon^{-1})$, respectively, and the same small communication complexity $\mathcal{O}(\epsilon^{-1}\ln\epsilon^{-1})$. Numerical experiments demonstrate that the proposed algorithms achieve lower sample and communication complexities than the existing decentralized AC algorithm. | reject |
Title: Manifold Mixup: Learning Better Representations by Interpolating Hidden States. Abstract: Deep networks often perform well on the data distribution on which they are trained, yet give incorrect (and often very confident) answers when evaluated on points from off of the training distribution. This is exemplified by the adversarial examples phenomenon but can also be seen in terms of model generalization and domain shift. Ideally, a model would assign lower confidence to points unlike those from the training distribution. We propose a regularizer which addresses this issue by training with interpolated hidden states and encouraging the classifier to be less confident at these points. Because the hidden states are learned, this has an important effect of encouraging the hidden states for a class to be concentrated in such a way so that interpolations within the same class or between two different classes do not intersect with the real data points from other classes. This has a major advantage in that it avoids the underfitting which can result from interpolating in the input space. We prove that the exact condition for this problem of underfitting to be avoided by Manifold Mixup is that the dimensionality of the hidden states exceeds the number of classes, which is often the case in practice. Additionally, this concentration can be seen as making the features in earlier layers more discriminative. We show that despite requiring no significant additional computation, Manifold Mixup achieves large improvements over strong baselines in supervised learning, robustness to single-step adversarial attacks, semi-supervised learning, and Negative Log-Likelihood on held out samples. | reject |
Title: Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processes. Abstract: There is a previously identified equivalence between wide fully connected neural networks (FCNs) and Gaussian processes (GPs). This equivalence enables, for instance, test set predictions that would have resulted from a fully Bayesian, infinitely wide trained FCN to be computed without ever instantiating the FCN, but by instead evaluating the corresponding GP. In this work, we derive an analogous equivalence for multi-layer convolutional neural networks (CNNs) both with and without pooling layers, and achieve state of the art results on CIFAR10 for GPs without trainable kernels. We also introduce a Monte Carlo method to estimate the GP corresponding to a given neural network architecture, even in cases where the analytic form has too many terms to be computationally feasible.
Surprisingly, in the absence of pooling layers, the GPs corresponding to CNNs with and without weight sharing are identical. As a consequence, translation equivariance, beneficial in finite channel CNNs trained with stochastic gradient descent (SGD), is guaranteed to play no role in the Bayesian treatment of the infinite channel limit - a qualitative difference between the two regimes that is not present in the FCN case. We confirm experimentally, that while in some scenarios the performance of SGD-trained finite CNNs approaches that of the corresponding GPs as the channel count increases, with careful tuning SGD-trained CNNs can significantly outperform their corresponding GPs, suggesting advantages from SGD training compared to fully Bayesian parameter estimation. | accept |
Title: HR-TD: A Regularized TD Method to Avoid Over-Generalization. Abstract: Temporal Difference learning with function approximation has been widely used recently and has led to several successful results. However, compared with the original tabular-based methods, one major drawback of temporal difference learning with neural networks and other function approximators is that they tend to over-generalize across temporally successive states, resulting in slow convergence and even instability. In this work, we propose a novel TD learning method, Hadamard product Regularized TD (HR-TD), that reduces over-generalization and thus leads to faster convergence. This approach can be easily applied to both linear and nonlinear function approximators.
HR-TD is evaluated on several linear and nonlinear benchmark domains, where we show improvement in learning behavior and performance. | reject |
Title: On the Spectral Bias of Neural Networks. Abstract: Neural networks are known to be a class of highly expressive functions able to fit even random input-output mappings with 100% accuracy. In this work we present properties of neural networks that complement this aspect of expressivity. By using tools from Fourier analysis, we show that deep ReLU networks are biased towards low frequency functions, meaning that they cannot have local fluctuations without affecting their global behavior. Intuitively, this property is in line with the observation that over-parameterized networks find simple patterns that generalize across data samples. We also investigate how the shape of the data manifold affects expressivity by showing evidence that learning high frequencies gets easier with increasing manifold complexity, and present a theoretical understanding of this behavior. Finally, we study the robustness of the frequency components with respect to parameter perturbation, to develop the intuition that the parameters must be finely tuned to express high frequency functions. | reject |
Title: Learning representations from temporally smooth data. Abstract: Events in the real world are correlated across nearby points in time, and we must learn from this temporally “smooth” data. However, when neural networks are trained to categorize or reconstruct single items, the common practice is to randomize the order of training items. What are the effects of temporally smooth training data on the efficiency of learning? We first tested the effects of smoothness in training data on incremental learning in feedforward nets and found that smoother data slowed learning. Moreover, sampling so as to minimize temporal smoothness produced more efficient learning than sampling randomly. If smoothness generally impairs incremental learning, then how can networks be modified to benefit from smoothness in the training data? We hypothesized that two simple brain-inspired mechanisms -- leaky memory in activation units and memory-gating -- could enable networks to exploit the redundancies in smooth data. Across all levels of data smoothness, these brain-inspired architectures achieved more efficient category learning than feedforward networks. Finally, we investigated how these brain-inspired mechanisms altered the internal representations learned by the networks. We found that networks with multi-scale leaky memory and memory-gating could learn internal representations that “un-mixed” data sources which vary on fast and slow timescales across training samples. Altogether, we identified simple mechanisms enabling neural networks to learn more quickly from temporally smooth data, and to generate internal representations that separate timescales in the training signal. | reject |
Title: Unsupervised Neural Multi-Document Abstractive Summarization of Reviews. Abstract: Abstractive summarization has been studied using neural sequence transduction methods with datasets of large, paired document-summary examples. However, such datasets are rare and the models trained from them do not generalize to other domains. Recently, some progress has been made in learning sequence-to-sequence mappings with only unpaired examples. In our work, we consider the setting where there are only documents (product or business reviews) with no summaries provided, and propose an end-to-end, neural model architecture to perform unsupervised abstractive summarization. Our proposed model consists of an auto-encoder trained so that the mean of the representations of the input reviews decodes to a reasonable summary-review. We consider variants of the proposed architecture and perform an ablation study to show the importance of specific components. We show through metrics and human evaluation that the generated summaries are highly abstractive, fluent, relevant, and representative of the average sentiment of the input reviews. | reject |