ICLR 2018 论文列表
6th International Conference on Learning Representations, ICLR 2018, Vancouver, BC, Canada, April 30 - May 3, 2018, Conference Track Proceedings.
|
On the Information Bottleneck Theory of Deep Learning.
Memory Architectures in Recurrent Neural Network Language Models.
Mixed Precision Training of Convolutional Neural Networks using Integer Operations.
Gaussian Process Behaviour in Wide Deep Neural Networks.
Variational Continual Learning.
Learning Sparse Neural Networks through L_0 Regularization.
Learning From Noisy Singly-labeled Data.
Sobolev GAN.
Training GANs with Optimism.
Variational Network Quantization.
Temporally Efficient Deep Learning with Spikes.
Deep Sensing: Active Sensing using Multi-directional Recurrent Neural Networks.
Multi-Mention Learning for Reading Comprehension with Neural Cascades.
Fix your classifier: the marginal value of training the last weight layer.
A New Method of Region Embedding for Text Classification.
A Compressed Sensing View of Unsupervised Text Embeddings, Bag-of-n-Grams, and LSTMs.
Divide-and-Conquer Reinforcement Learning.
Hierarchical and Interpretable Skill Acquisition in Multi-task Reinforcement Learning.
N2N learning: Network to Network Compression via Policy Gradient Reinforcement Learning.
Progressive Reinforcement Learning with Distillation for Multi-Skilled Motion Control.
Memory Augmented Control Networks.
Overcoming Catastrophic Interference using Conceptor-Aided Backpropagation.
Active Neural Localization.
Neural Map: Structured Memory for Deep Reinforcement Learning.
Eigenoption Discovery through the Deep Successor Representation.
On the regularization of Wasserstein GANs.
Robustness of Classifiers to Universal Perturbations: A Geometric Perspective.
Stochastic gradient descent performs variational inference, converges to limit cycles for deep networks.
Online Learning Rate Adaptation with Hypergradient Descent.
When is a Convolutional Filter Easy to Learn?
Policy Optimization by Genetic Distillation.
Guide Actor-Critic for Continuous Control.
Boosting the Actor with Dual Critic.
Adaptive Quantization of Neural Networks.
Residual Loss Prediction: Reinforcement Learning With No Incremental Feedback.
Alternating Multi-bit Quantization for Recurrent Neural Networks.
TreeQN and ATreeC: Differentiable Tree-Structured Models for Deep Reinforcement Learning.
Temporal Difference Models: Model-Free Deep RL for Model-Based Control.
DORA The Explorer: Directed Outreaching Reinforcement Action-Selection.
TD or not TD: Analyzing the Role of Temporal Differencing in Deep Reinforcement Learning.
mixup: Beyond Empirical Risk Minimization.
Deep Voice 3: Scaling Text-to-Speech with Convolutional Sequence Learning.
Non-Autoregressive Neural Machine Translation.
PixelNN: Example-based Image Synthesis.
Learning to Teach.
Auto-Encoding Sequential Monte Carlo.
Synthesizing realistic neural population activity patterns using Generative Adversarial Networks.
Parameter Space Noise for Exploration.
SMASH: One-Shot Model Architecture Search through HyperNetworks.
An image representation based convolutional network for DNA classification.
Improving the Universality and Learnability of Neural Programmer-Interpreters with Combinator Abstraction.
Expressive power of recurrent neural networks.
Towards Synthesizing Complex Programs From Input-Output Examples.
Deep Learning and Quantum Entanglement: Fundamental Connections with Implications to Network Design.
A Simple Neural Attentive Meta-Learner.
Learning Robust Rewards with Adverserial Inverse Reinforcement Learning.
Learning to Multi-Task by Active Sampling.
Gradient Estimators for Implicit Models.
Self-ensembling for visual domain adaptation.
Understanding Short-Horizon Bias in Stochastic Meta-Optimization.
WHAI: Weibull Hybrid Autoencoding Inference for Deep Topic Modeling.
Learning Sparse Latent Representations with the Deep Copula Information Bottleneck.
Boundary Seeking GANs.
Learning a Generative Model for Validity in Complex Discrete Structures.
Debiasing Evidence Approximations: On Importance-weighted Autoencoders and Jackknife Variational Inference.
On Unifying Deep Generative Models.
Backpropagation through the Void: Optimizing control variates for black-box gradient estimation.
Learning Awareness Models.
Understanding image motion with group representations.
Predicting Floor-Level for 911 Calls with Neural Networks and Smartphone Sensor Data.
Spatially Transformed Adversarial Examples.
Generating Natural Adversarial Examples.
Deep Gaussian Embedding of Graphs: Unsupervised Inductive Learning via Ranking.
Detecting Statistical Interactions from Neural Network Weights.
Learning how to explain neural networks: PatternNet and PatternAttribution.
Not-So-Random Features.
SpectralNet: Spectral Clustering using Deep Neural Networks.
Global Optimality Conditions for Deep Neural Networks.
Loss-aware Weight Quantization of Deep Networks.
Active Learning for Convolutional Neural Networks: A Core-Set Approach.
Scalable Private Learning with PATE.
Combining Symbolic Expressions and Black-box Function Evaluations in Neural Programs.
Reinforcement Learning on Web Interfaces using Workflow-Guided Exploration.
Hierarchical Representations for Efficient Architecture Search.
Beyond Shared Hierarchies: Deep Multitask Learning through Soft Layer Ordering.
Compositional Attention Networks for Machine Reasoning.
Dynamic Neural Program Embeddings for Program Repair.
The Role of Minimal Complexity Functions in Unsupervised Learning of Semantic Mappings.
Lifelong Learning with Dynamically Expandable Networks.
Go for a Walk and Arrive at the Answer: Reasoning Over Paths in Knowledge Bases using Reinforcement Learning.
Deep Active Learning for Named Entity Recognition.
Learning Intrinsic Sparse Structures within Long Short-Term Memory.
Neural Language Modeling by Jointly Learning Syntax and Lexicon.
FusionNet: Fusing via Fully-aware Attention with Application to Machine Comprehension.
Improving the Improved Training of Wasserstein GANs: A Consistency Term and Its Dual Effect.
Coulomb GANs: Provably Optimal Nash Equilibria via Potential Fields.
Activation Maximization Generative Adversarial Nets.
Training Generative Adversarial Networks via Primal-Dual subgradient Methods: a Lagrangian Perspective on GaN.
Learning Wasserstein Embeddings.
CausalGAN: Learning Causal Implicit Generative Models with Adversarial Training.
Matrix capsules with EM routing.
Decision Boundary Analysis of Adversarial Examples.
Mitigating Adversarial Effects Through Randomization.
Cascade Adversarial Machine Learning Regularized with a Unified Embedding.
Can Neural Networks Understand Logical Entailment?
Consequentialist conditional cooperation in social dilemmas with imperfect information.
Leave no Trace: Learning to Reset for Safe and Autonomous Reinforcement Learning.
Reinforcement Learning Algorithm Selection.
Intrinsic Motivation and Automatic Curricula via Asymmetric Self-Play.
Distributed Fine-tuning of Language Models on Private Data.
Multi-Task Learning for Document Ranking and Query Suggestion.
Natural Language Inference over Interaction Space.
Learning General Purpose Distributed Sentence Representations via Large Scale Multi-task Learning.
All-but-the-Top: Simple and Effective Postprocessing for Word Representations.
Word translation without parallel data.
DCN+: Mixed Objective And Deep Residual Coattention for Question Answering.
Regularizing and Optimizing LSTM Language Models.
Sensitivity and Generalization in Neural Networks: an Empirical Study.
Implicit Causal Models for Genome-wide Association Studies.
A Bayesian Perspective on Generalization and Stochastic Gradient Descent.
Adaptive Dropout with Rademacher Complexity Regularization.
Many Paths to Equilibrium: GANs Do Not Need to Decrease a Divergence At Every Step.
The Implicit Bias of Gradient Descent on Separable Data.
On the importance of single directions for generalization.
A PAC-Bayesian Approach to Spectrally-Normalized Margin Bounds for Neural Networks.
SGD Learns Over-parameterized Networks that Provably Generalize on Linearly Separable Data.
Neumann Optimizer: A Practical Optimization Algorithm for Deep Neural Networks.
Proximal Backpropagation.
Kronecker-factored Curvature Approximations for Recurrent Neural Networks.
Don't Decay the Learning Rate, Increase the Batch Size.
Recasting Gradient-Based Meta-Learning as Hierarchical Bayes.
Monotonic Chunkwise Attention.
Learn to Pay Attention.
Training wide residual networks for deployment using a single bit for each weight.
Understanding Deep Neural Networks with Rectified Linear Units.
Towards Reverse-Engineering Black-Box Neural Networks.
Do GANs learn the distribution? Some Theory and Empirics.
FearNet: Brain-Inspired Model for Incremental Learning.
Wavelet Pooling for Convolutional Neural Networks.
Routing Networks: Adaptive Selection of Non-Linear Functions for Multi-Task Learning.
Bi-Directional Block Self-Attention for Fast and Memory-Efficient Sequence Modeling.
Skip Connections Eliminate Singularities.
Deep Complex Networks.
Learning to cluster in order to transfer across domains and tasks.
Generalizing Across Domains via Cross-Gradient Training.
A DIRT-T Approach to Unsupervised Domain Adaptation.
Meta-Learning for Semi-Supervised Few-Shot Classification.
A Framework for the Quantitative Evaluation of Disentangled Representations.
Semantically Decomposing the Latent Spaces of Generative Adversarial Networks.
Few-Shot Learning with Graph Neural Networks.
Learning a neural response metric for retinal prosthesis.
Emergence of grid-like representations by training recurrent neural networks to perform spatial localization.
Identifying Analogies Across Domains.
Hierarchical Density Order Embeddings.
SCAN: Learning Hierarchical Compositional Visual Concepts.
Compositional Obverter Communication Learning from Raw Visual Input.
Few-shot Autoregressive Density Estimation: Towards Learning to Learn Distributions.
Generative Models of Visually Grounded Imagination.
Relational Neural Expectation Maximization: Unsupervised Discovery of Objects and their Interactions.
Simulated+Unsupervised Learning With Adaptive Data Generation and Bidirectional Mappings.
Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting.
Rethinking the Smaller-Norm-Less-Informative Assumption in Channel Pruning of Convolution Layers.
On the Expressive Power of Overlapping Architectures of Deep Learning.
Critical Percolation as a Framework to Analyze the Training of Deep Networks.
Generative networks as inverse problems with Scattering transforms.
Improving GAN Training via Binarized Representation Entropy (BRE) Regularization.
Quantitatively Evaluating GANs With Divergences Proposed for Training.
Deep Rewiring: Training very sparse deep networks.
Learning Discrete Weights Using the Local Reparameterization Trick.
Deep Autoencoding Gaussian Mixture Model for Unsupervised Anomaly Detection.
A Hierarchical Model for Device Placement.
Noisy Networks For Exploration.
Depthwise Separable Convolutions for Neural Machine Translation.
Attacking Binarized Neural Networks.
Parallelizing Linear Recurrent Neural Nets Over Sequence Length.
Can recurrent neural networks warp time?
Fraternal Dropout.
Ensemble Adversarial Training: Attacks and Defenses.
Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Models.
Certified Defenses against Adversarial Examples.
PixelDefend: Leveraging Generative Models to Understand and Defend against Adversarial Examples.
Initialization matters: Orthogonal Predictive State Recurrent Neural Networks.
Memory-based Parameter Adaptation.
On the State of the Art of Evaluation in Neural Language Models.
Towards Neural Phrase-based Machine Translation.
Multi-level Residual Networks from Dynamical Systems View.
Neural Speed Reading via Skim-RNN.
Unsupervised Cipher Cracking Using Discrete GANs.
Simulating Action Dynamics with Neural Process Networks.
Communication Algorithms via Deep Learning.
Deep Learning for Physical Processes: Incorporating Prior Scientific Knowledge.
Towards Deep Learning Models Resistant to Adversarial Attacks.
HexaConv.
Evaluating the Robustness of Neural Networks: An Extreme Value Theory Approach.
i-RevNet: Deep Invertible Networks.
Learning to Count Objects in Natural Images for Visual Question Answering.
Semi-parametric topological memory for navigation.
Emergent Communication through Negotiation.
Residual Connections Encourage Iterative Inference.
Universal Agent for Disentangling Environments and Tasks.
Emergent Complexity via Multi-Agent Competition.
Interactive Grounded Language Acquisition and Generalization in a 2D World.
Interpretable Counting for Visual Question Answering.
Twin Networks: Matching the Future for Sequence Generation.
Modular Continual Learning in a Unified Visual Environment.
Skip RNN: Learning to Skip State Updates in Recurrent Neural Networks.
Countering Adversarial Images using Input Transformations.
Towards better understanding of gradient-based attribution methods for Deep Neural Networks.
Automatically Inferring Data Quality for Spatiotemporal Forecasting.
Towards Image Understanding from Deep Compression Without Decoding.
Stochastic Variational Video Prediction.
Trust-PCL: An Off-Policy Trust Region Method for Continuous Control.
Thermometer Encoding: One Hot Way To Resist Adversarial Examples.
GANITE: Estimation of Individualized Treatment Effects using Generative Adversarial Nets.
Sparse Persistent RNNs: Squeezing Large Recurrent Networks On-Chip.
Stochastic Activation Pruning for Robust Adversarial Defense.
Memorization Precedes Generation: Learning Unsupervised GANs with Memory Networks.
Measuring the Intrinsic Dimension of Objective Landscapes.
Unbiased Online Recurrent Optimization.
Decision-Based Adversarial Attacks: Reliable Attacks Against Black-Box Machine Learning Models.
On the Discrimination-Generalization Tradeoff in GANs.
Empirical Risk Landscape Analysis for Understanding Deep Neural Networks.
The power of deeper networks for expressing natural functions.
Learning Parametric Closed-Loop Policies for Markov Potential Games.
Critical Points of Linear Neural Networks: Analytical Forms and Landscape Properties.
Learning One-hidden-layer Neural Networks with Landscape Design.
Unsupervised Neural Machine Translation.
QANet: Combining Local Convolution with Global Self-Attention for Reading Comprehension.
Deep Gradient Compression: Reducing the Communication Bandwidth for Distributed Training.
Compressing Word Embeddings via Deep Compositional Code Learning.
A Deep Reinforced Model for Abstractive Summarization.
Unsupervised Machine Translation Using Monolingual Corpora Only.
Generating Wikipedia by Summarizing Long Sequences.
Mastering the Dungeon: Grounded Language Learning by Mechanical Turker Descent.
Learning Differentially Private Recurrent Language Models.
Large scale distributed neural network training through online distillation.
VoiceLoop: Voice Fitting and Synthesis via a Phonological Loop.
Training Confidence-calibrated Classifiers for Detecting Out-of-Distribution Samples.
Learning from Between-class Examples for Deep Sound Recognition.
Distributed Prioritized Experience Replay.
Apprentice: Using Knowledge Distillation Techniques To Improve Low-Precision Network Accuracy.
The High-Dimensional Geometry of Binary Neural Networks.
A Scalable Laplace Approximation for Neural Networks.
Kernel Implicit Variational Inference.
Flipout: Efficient Pseudo-Independent Weight Perturbations on Mini-Batches.
Variational Inference of Disentangled Latent Concepts from Unlabeled Observations.
Variational image compression with a scale hyperprior.
Action-dependent Control Variates for Policy Optimization via Stein Identity.
Variational Message Passing with Structured Inference Networks.
Model compression via distillation and quantization.
Learning to Share: simultaneous parameter tying and Sparsification in Deep Learning.
Learning Approximate Inference Networks for Structured Prediction.
Deep Learning as a Mixed Convex-Combinatorial Optimization Problem.
Smooth Loss Functions for Deep Top-k Classification.
Demystifying MMD GANs.
Adversarial Dropout Regularization.
Learning Latent Representations in Neural Networks for Clustering through Pseudo Supervision and Graph-based Activity Regularization.
NerveNet: Learning Structured Policy with Graph Neural Networks.
An efficient framework for learning sentence representations.
Emergent Translation in Multi-Agent Communication.
FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling.
Emergent Communication in a Multi-Modal, Multi-Step Referential Game.
Unsupervised Representation Learning by Predicting Image Rotations.
cGANs with Projection Discriminator.
Viterbi-based Pruning for Sparse Matrix with Fixed and High Index Compression Ratio.
Parametrized Hierarchical Procedures for Neural Programming.
Hierarchical Subtask Discovery with Non-Negative Matrix Factorization.
Distributed Distributional Deterministic Policy Gradients.
SEARNN: Training RNNs with global-local losses.
The Reactor: A fast and sample-efficient Actor-Critic agent for Reinforcement Learning.
MGAN: Training Generative Adversarial Nets with Multiple Generators.
WRPN: Wide Reduced-Precision Networks.
Evidence Aggregation for Answer Re-Ranking in Open-Domain Question Answering.
Neural-Guided Deductive Search for Real-Time Program Synthesis from Examples.
Syntax-Directed Variational Autoencoder for Structured Data.
Deep Neural Networks as Gaussian Processes.
Meta Learning Shared Hierarchies.
Maximum a Posteriori Policy Optimisation.
Meta-Learning and Universality: Deep Representations and Gradient Descent can Approximate any Learning Algorithm.
Divide and Conquer Networks.
MaskGAN: Better Text Generation via Filling in the _______.
Latent Constraints: Learning to Generate Conditionally from Unconditional Generative Models.
Mixed Precision Training.
The Kanerva Machine: A Generative Distributed Memory.
Improving GANs Using Optimal Transport.
An Online Learning Approach to Generative Adversarial Networks.
Generalizing Hamiltonian Monte Carlo with Neural Networks.
Minimax Curriculum Learning: Machine Teaching with Desirable Difficulties and Scheduled Diversity.
Graph Attention Networks.
Stabilizing Adversarial Nets with Prediction Methods.
Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks.
Polar Transformer Networks.
Decoupling the Layers in Residual Networks.
Auto-Conditioned Recurrent Networks for Extended Complex Human Motion Synthesis.
Espresso: Efficient Forward Propagation for Binary Deep Neural Networks.
Efficient Sparse-Winograd Convolutional Neural Networks.
Leveraging Grammar and Reinforcement Learning for Neural Program Synthesis.
Hyperparameter optimization: a spectral approach.
Imitation Learning from Visual Data with Multiple Intentions.
Latent Space Oddity: on the Curvature of Deep Generative Models.
Fidelity-Weighted Learning.
Semantic Interpolation in Implicit Models.
Deep Bayesian Bandits Showdown: An Empirical Comparison of Bayesian Deep Networks for Thompson Sampling.
Multi-View Data Generation Without View Supervision.
Unsupervised Learning of Goal Spaces for Intrinsically Motivated Goal Exploration.
Learning an Embedding Space for Transferable Robot Skills.
Learning Latent Permutations with Gumbel-Sinkhorn Networks.
Deep Learning with Logged Bandit Feedback.
A Neural Representation of Sketch Drawings.
Model-Ensemble Trust-Region Policy Optimization.
Truncated horizon Policy Search: Combining Reinforcement Learning & Imitation Learning.
Large Scale Optimal Transport and Mapping Estimation.
Minimal-Entropy Correlation Alignment for Unsupervised Deep Domain Adaptation.
AmbientGAN: Generative models from lossy measurements.
Beyond Word Importance: Contextual Decomposition to Extract Interactions from LSTMs.
Zero-Shot Visual Imitation.
Variance Reduction for Policy Gradient with Action-Dependent Factorized Baselines.
Progressive Growing of GANs for Improved Quality, Stability, and Variation.
Neural Sketch Learning for Conditional Program Generation.
Boosting Dilated Convolutional Networks with Mixed Tensor Decompositions.
Continuous Adaptation via Meta-Learning in Nonstationary and Competitive Environments.
Breaking the Softmax Bottleneck: A High-Rank RNN Language Model.
Characterizing Adversarial Subspaces Using Local Intrinsic Dimensionality.
Learning to Represent Programs with Graphs.
Spectral Normalization for Generative Adversarial Networks.
Wasserstein Auto-Encoders.
Learning Deep Mean Field Games for Modeling Large Population Behavior.
Certifying Some Distributional Robustness with Principled Adversarial Training.
On the insufficiency of existing momentum schemes for Stochastic Optimization.
Ask the Right Questions: Active Question Reformulation with Reinforcement Learning.
Spherical CNNs.
Emergence of Linguistic Communication from Referential Games with Symbolic and Pixel Input.
Training and Inference with Integers in Deep Neural Networks.
Multi-Scale Dense Networks for Resource Efficient Image Classification.
Synthetic and Natural Noise Both Break Neural Machine Translation.
On the Convergence of Adam and Beyond.
