【重磅】深度学习顶会 ICLR 2018 匿名提交论文列表（附pdf下载链接）
点击上方“专知”关注获取更多AI知识!
【导读】ICLR，全称为「International Conference on Learning Representations」（国际学习表征会议），2013 年才刚刚成立了第一届。这个一年一度的会议虽然今年2017年办到第六届，已经被学术研究者们广泛认可，被认为「深度学习的顶级会议」。这个会议由位列深度学习三大巨头之二的 Yoshua Bengio 和 Yann LeCun 牵头创办。Yoshua Bengio 是蒙特利尔大学教授，深度学习三巨头之一，他领导蒙特利尔大学的人工智能实验室（MILA）进行 AI 技术的学术研究。MILA 是世界上最大的人工智能研究中心之一，与谷歌也有着密切的合作。 Yann LeCun 就自不用提，同为深度学习三巨头之一的他现任 Facebook 人工智能研究院（FAIR）院长、纽约大学教授。作为卷积神经网络之父，他为深度学习的发展和创新作出了重要贡献。
ICLR 采用Open Review 评审制度。Open Review 则非常不同，根据规定，所有提交的论文都会公开姓名等信息，并且接受所有同行的评价及提问（open peer review），任何学者都可或匿名或实名地评价论文。而在公开评审结束后，论文作者也能够对论文进行调整和修改。目前 ICLR 的历届所有论文及评审讨论的内容，都完整地保存在 OpenReview.net 上，它也是 ICLR 的官方投稿入口。OpenReview.net 是马萨诸塞大学阿默斯特学院 Andrew McCallum 为 ICLR 2013 牵头创办的一个公开评审系统，秉承公开同行评审、公开发表、公开来源、公开讨论、公开引导、公开推荐、公开 API 及开源等八大原则，得到了 Facebook、Google、NSF 和马萨诸塞大学阿默斯特中心等机构的支持。
以下为论文列表：
来源：https://openreview.net/group?id=ICLR.cc/2018/Conference
专知进行关键词统计信息如下：
可以看出 深度学习 神经网络 生成式对抗网络、强化学习、循环神经网络等等是投稿论文热点。
论文列表：
《Improving Discriminator-Generator Balance in Generative Adversarial Networks》:
下载地址：https://openreview.net/pdf/b9ca5077f6a0c9481b172ad051d0bff48f2949c2.pdf
《Placeholder》:
下载地址：https://openreview.net/pdf/a3ee124c0cc5f02acc976ae67f563ea632fbe23d.pdf
《Complex- and Real-Valued Neural Network Architectures》:
关键词：complex numbers complex-valued neural network multi-layer perceptron architecture
下载地址：https://openreview.net/pdf/4127a6a37a17384ef2d001931450550a33b69acd.pdf
《Revisiting Knowledge Base Embedding as Tensor Decomposition》:
关键词：Knowledge base embedding
下载地址：https://openreview.net/pdf/4e9e3d851b60e8aa75b53c344e0ed3988c5300fa.pdf
《Tree2Tree Learning with Memory Unit》:
下载地址：https://openreview.net/pdf/9558215bb47a09abcef80ac65b52474a09da0be1.pdf
《Combining Model-based and Model-free RL via Multi-step Control Variates》:
下载地址：https://openreview.net/pdf/c94761f85f8bdbd8b9c53261e25b4ec0258406e8.pdf
《Hyperedge2vec: Distributed Representations for Hyperedges》:
关键词：hypergraph representation learning tensors
下载地址：https://openreview.net/pdf/53c0248eb3e4d4fff5dd84d97ce5132f5d5861bf.pdf
《Deep Complex Networks》:
关键词：deep learning complex-valued neural networks
下载地址：https://openreview.net/pdf/21bc670e37fcb28f944d33f287f626306b316875.pdf
《OMIE: The Online Mutual Information Estimator》:
关键词：Deep Learning Neural Networks Information Theory Generative models
下载地址：https://openreview.net/pdf/0d736ada7e156b950fdd5eb287d9f95a22d9c54c.pdf
《Few-Shot Learning with Variational Homoencoders》:
关键词：generative models one-shot learning metalearning pixelcnn hierarchical bayesian omniglot
下载地址：https://openreview.net/pdf/36668c5f207557f4d40dcb81393774d2f0908266.pdf
《Video Action Segmentation with Hybrid Temporal Networks》:
关键词：action segmentation video labeling temporal networks
下载地址：https://openreview.net/pdf/9cb1db4642c01584e6ca3c886e730f3743542a24.pdf
《Learning Efficient Tensor Representations with Ring Structure Networks》:
关键词：Tensor Decomposition Tensor Networks Stochastic Gradient Descent
下载地址：https://openreview.net/pdf/a2f569c8fabb4aa65611d077829bfff2946df00d.pdf
《Fitting Data Noise in Variational Autoencoder》:
关键词：variational autoencoder noise modelling representation learning generative model disentanglement
下载地址：https://openreview.net/pdf/62a904438b7296e9a4a604381c06ee828574d98b.pdf
《Bayesian Uncertainty Estimation for Batch Normalized Deep Networks》:
关键词：uncertainty estimation deep learning Bayesian learning batch normalization
下载地址：https://openreview.net/pdf/ac74faafa0bba2c7808c4d9991b7b711ee064038.pdf
《A Goal-oriented Neural Conversation Model by Self-Play》:
关键词：conversation model seq2seq self-play reinforcement learning
下载地址：https://openreview.net/pdf/40fc8cdd76f4aba7cb8069509d9e5ddf2523ad35.pdf
《Automatic Goal Generation for Reinforcement Learning Agents》:
关键词：Reinforcement Learning Multi-task Learning Curriculum Learning
下载地址：https://openreview.net/pdf/7efb4d89e4f175f4cdecd4783b8b5a5d8af797cf.pdf
《A novel method to determine the number of latent dimensions with SVD》:
关键词：SVD Latent Dimensions Dimension Reductions Machine Learning
下载地址：https://openreview.net/pdf/5a5d920c9b7b9b39015b595683426873a38b3e8b.pdf
《Universal Agent for Disentangling Environments and Tasks》:
关键词：reinforcement learning transfer learning
下载地址：https://openreview.net/pdf/d18b693b43b8866425b41c5a3ae6e4de9b45658d.pdf
《Covariant Compositional Networks For Learning Graphs》:
关键词：graph neural networks message passing label propagation equivariant representation
下载地址：https://openreview.net/pdf/7673e6cf0b07d195633b82c9905e205759f686e9.pdf
《Deep learning mutation prediction enables early stage lung cancer detection in liquid biopsy》:
关键词：somatic mutation variant calling cancer liquid biopsy early detection convolution deep learning machine learning lung cancer error suppression mutect
下载地址：https://openreview.net/pdf/3da2a17bf6bec5ff1a8f0dd52c100ceb17694e76.pdf
《Learning To Generate Reviews and Discovering Sentiment》:
关键词：unsupervised learning representation learning deep learning
下载地址：https://openreview.net/pdf/82eaeeca82af695721cc73403066982e93ef60d2.pdf
《Noise-Based Regularizers for Recurrent Neural Networks》:
下载地址：https://openreview.net/pdf/f5434c16d9149ba2ecf5dff8e5b5a34dce8e600b.pdf
《Prediction Under Uncertainty with Error Encoding Networks》:
下载地址：https://openreview.net/pdf/bd3b0e1996f51903fe07077607eeae4c2b1bbafd.pdf
《Genative Entity Networks: Disentangling Entitites and Attributes in Visual Scenes using Partial Natural Language Descriptions》:
关键词：VAE Generative Model Vision Natural Language
下载地址：https://openreview.net/pdf/3cf45610469af5c3ecdef0638ed8c83937f59c27.pdf
《WSNet: Learning Compact and Efficient Networks with Weight Sampling》:
关键词：Deep learning model compression
下载地址：https://openreview.net/pdf/53e7e6f6b94dca95f61fbed0fcaf988215ad2083.pdf
《TD Learning with Constrained Gradients》:
关键词：Reinforcement Learning TD Learning DQN
下载地址：https://openreview.net/pdf/424ef3a312b7502cf11a36f4693095fb81db7ecb.pdf
《Improving the Improved Training of Wasserstein GANs》:
关键词：GAN WGAN
下载地址：https://openreview.net/pdf/98bba828944f13faf32019e9400c7ce9615e175e.pdf
《Exploring Representation Methods for Sequence Labeling》:
下载地址：https://openreview.net/pdf/efa84800de59a703122ea1f328a6a3c1031e1cfa.pdf
《Fraternal Dropout》:
关键词：fraternal dropout activity regularization recurrent neural networks RNN LSTM faster convergence
下载地址：https://openreview.net/pdf/e58a67feb2152ae4cd53042cbb8762df63757b73.pdf
《What are image captions made of?》:
关键词：image captioning representation learning interpretability rnn multimodal vision to language
下载地址：https://openreview.net/pdf/0e647e0120fb1714b378c172dbf1934d6c901237.pdf
《Sequential Coordination of Deep Models for Learning Visual Arithmetic》:
关键词：reinforcement learning pretrained deep learning perception algorithmic
下载地址：https://openreview.net/pdf/3aabac9a13b73eaca48e53acec3f071ba9fb96b9.pdf
《DETECTING ADVERSARIAL PERTURBATIONS WITH SALIENCY》:
关键词：Adversarial Examples Detection Saliency Model Interpretation
下载地址：https://openreview.net/pdf/b7aafb6a6dbb956dea1e53cf9f4a58ec39e9513b.pdf
《An inference-based policy gradient method for learning options》:
关键词：reinforcement learning hierarchy options inference
下载地址：https://openreview.net/pdf/7ff2f7d7dba366ae35b85d4dbac7d2a46c59007e.pdf
《Generative Entity Networks: Disentangling Entities and Attributes in Visual Scenes using Partial Natural Language Descriptions》:
关键词：VAE Vision Natural Language
下载地址：https://openreview.net/pdf/bfd58631af339d8043d30210ba8c2ad9d965cc3e.pdf
《Don’t encrypt the data; just approximate the model \ Towards Secure Transaction and Fair Pricing of Training Data》:
关键词：Applications Security in Machine Learning Fairness and Security Model Compression
下载地址：https://openreview.net/pdf/69170f53ffe9f431f2c54cd1a453add292d356cb.pdf
《Alpha-divergence bridges maximum likelihood and reinforcement learning in neural sequence generation》:
关键词：neural network reinforcement learning natural language processing machine translation alpha-divergence
下载地址：https://openreview.net/pdf/4122d80b6740caf9641d8bbc9dc1cf00e2259f51.pdf
《3C-GAN: AN CONDITION-CONTEXT-COMPOSITE GENERATIVE ADVERSARIAL NETWORKS FOR GENERATING IMAGES SEPARATELY》:
下载地址：https://openreview.net/pdf/554e41c5738f9a1f35ea2eae5a31bebad2354fe6.pdf
《Parametric Information Bottleneck to \Optimize Stochastic Neural Networks》:
关键词：Information Bottleneck Deep Neural Networks
下载地址：https://openreview.net/pdf/db367bd113d779803710f2c0b70e6a13fa0e508d.pdf
《Towards a Testable Notion of Generalization for Generative Adversarial Networks》:
关键词：generative adversarial networks Wasserstein GAN generalization theory
下载地址：https://openreview.net/pdf/c8e2421cd23954c4dc741562cc8192c356fd3068.pdf
《TOWARDS ROBOT VISION MODULE DEVELOPMENT WITH EXPERIENTIAL ROBOT LEARNING》:
关键词：Deep Learning Robotics Artificial Intelligence Computer Vision
下载地址：https://openreview.net/pdf/00e5c4aefc80d0396ee745c032d27e0bccb43079.pdf
《Variational Bi-LSTMs》:
下载地址：https://openreview.net/pdf/4324fa39868648281fcca9536b21bab92f264995.pdf
《Learning an Embedding Space for Transferable Robot Skills》:
关键词：Deep Reinforcement Learning Variational Inference Control Robotics
下载地址：https://openreview.net/pdf/91cf23f41853ce25a71700dc007240032056772d.pdf
《ON MODELING HIERARCHICAL DATA VIA ENCAPSULATION OF PROBABILITY DENSITIES》:
关键词：embeddings
下载地址：https://openreview.net/pdf/a09f1ca6968a32ebc27f80d50c9cf7afcdeaaca5.pdf
《withdraw》:
下载地址：https://openreview.net/pdf/210160b60e7b9c27d7075e84fb18ad70b9641847.pdf
《Neural Compositional Denotational Semantics for Question Answering》:
关键词：question answering knowledge graph compositional model semantics
下载地址：https://openreview.net/pdf/576e30e63197e5c48e28f9a662cf7d1f7e0a7424.pdf
《Model compression via distillation and quantization》:
下载地址：https://openreview.net/pdf/6a770d7c95ac938be4c78c7d38abb92a01749769.pdf
《Binarized Back-Propagation: Training Binarized Neural Networks with Binarized Gradients》:
关键词：Neural Network acceleration Low Precision neural networks.
下载地址：https://openreview.net/pdf/8c46133b2c265d251eb6b79476877fd072e2445e.pdf
《DON’T ENCRYPT THE DATA, JUST APPROXIMATE THE MODEL/ TOWARDS SECURE TRANSACTION AND FAIR PRICING OF TRAINING DATA》:
关键词：Security in Machine Learning Information Security Fairness and Privacy
下载地址：https://openreview.net/pdf/6b168938dbf6014d12195848c4dc000920a179b3.pdf
《Optimal transport maps for distribution preserving operations on latent spaces of Generative Models》:
关键词：GANs transport
下载地址：https://openreview.net/pdf/b7c56e1cd66dbf15ef3b4bc4d2aa145c07b24d94.pdf
《Learning Representations for Faster Similarity Search》:
下载地址：https://openreview.net/pdf/6c1f3ff600aabd6e41f45bbef2b086a6595aea5a.pdf
《Maximum a Posteriori Policy Optimisation》:
关键词：Reinforcement Learning Variational Inference Control
下载地址：https://openreview.net/pdf/84a8906ab0166521e2bafc00d0b1a21a077f4f8d.pdf
《MaskGAN: Textual Generative Adversarial Networks from Filling-in-the-Blank》:
关键词：Deep learning GAN
下载地址：https://openreview.net/pdf/0f1ccb61544842755137c3af63c473b6b44b5948.pdf
《Do Convolutional Neural Networks act as Compositional Nearest Neighbors?》:
关键词：interpreting convolutional neural networks nearest neighbors generative adversarial networks
下载地址：https://openreview.net/pdf/a2fc659e794dc6e3cc0cb0d7598a70c1046fd8a9.pdf
《Kernel Implicit Variational Inference》:
下载地址：https://openreview.net/pdf/37257d67235ce949317db3948d11ee647fcb9743.pdf
《THINK VISUALLY: QUESTION ANSWERING THROUGH VIRTUAL IMAGERY》:
下载地址：https://openreview.net/pdf/37eda3658e01744d88c216dea06c26b4595eb965.pdf
《BLOCK-NORMALIZED GRADIENT METHOD: AN EMPIRICAL STUDY FOR TRAINING DEEP NEURAL NETWORK》:
下载地址：https://openreview.net/pdf/0f2afdd3f1854921fca69289a77b0a0b34922fe1.pdf
《Autonomous Vehicle Fleet Coordination With Deep Reinforcement Learning》:
关键词：Deep Reinforcement Learning mult-agent systems
下载地址：https://openreview.net/pdf/c7c8805df55c3c06682680157594fc6adcc1686c.pdf
《Kronecker-factored Curvature Approximations for Recurrent Neural Networks》:
关键词：optimization K-FAC natural gradient recurrent neural networks
下载地址：https://openreview.net/pdf/4067d0408e56145839859729549fb5ad062a9820.pdf
《POLICY DRIVEN GENERATIVE ADVERSARIAL NETWORKS FOR ACCENTED SPEECH GENERATION》:
关键词：speech generation accent gan adversarial reinforcement memory lstm policy gradients human
下载地址：https://openreview.net/pdf/cf54a1d15e4d0b075be4be9888e8810789404941.pdf
《Scalable Private Learning with PATE》:
关键词：privacy differential privacy machine learning deep learning
下载地址：https://openreview.net/pdf/85d216249a97e2df4929959351d82854c163e077.pdf
《AMPNet: Asynchronous Model-Parallel Training for Dynamic Neural Networks》:
关键词：asynchronous neural network deep learning graph tree rnn
下载地址：https://openreview.net/pdf/19a9932d9f73b3be2d71a8d7f335e6cb2e53236f.pdf
《Connectivity Learning in Multi-Branch Networks》:
关键词：connectivity learning multi-branch networks image categorization
下载地址：https://openreview.net/pdf/8d9ca4f54cbd879c63926ba3c6a466f8593c6168.pdf
《GATED FAST WEIGHTS FOR ASSOCIATIVE RETRIEVAL》:
关键词：fast weights RNN associative retrieval time-varying variables
下载地址：https://openreview.net/pdf/d184c06b932921a0bf6988238c0adc3e99c00d9a.pdf
《Generating Adversarial Examples with Adversarial Networks》:
关键词：adversarial examples generative adversarial network black-box attack
下载地址：https://openreview.net/pdf/e074cf9e1b09af73e658886fd664e257da48c396.pdf
《Online Learning Rate Adaptation with Hypergradient Descent》:
下载地址：https://openreview.net/pdf/99866a78dad33243f7634f3657ba2bbbc7825f79.pdf
《Relational Neural Expectation Maximization》:
关键词：Common-sense Physical Reasoning Intuitive Physics Representation Learning Model building
下载地址：https://openreview.net/pdf/0226604602a78f0f53c9a773e1888b4efedfb0e0.pdf
《Learning Awareness Models》:
关键词：Awareness Prediction Seq2seq Robots
下载地址：https://openreview.net/pdf/65f824e1d702c5e541a08c20c9700fd3c3c6aa8f.pdf
《Revisiting The Master-Slave Architecture In Multi-Agent Deep Reinforcement Learning》:
关键词：Deep Reinforcement Learning Multi-Agent Reinforcement Learning StarCraft Micromanagement Tasks
下载地址：https://openreview.net/pdf/9546b4f4eddb1629ead14f21afcee2ae6e9759b3.pdf
《STRUCTURED ALIGNMENT NETWORKS》:
关键词：structured attention sentence matching
下载地址：https://openreview.net/pdf/7cc678302888436c96fed050dfb89b10e20932d2.pdf
《On the regularization of Wasserstein GANs》:
下载地址：https://openreview.net/pdf/5322df3532cfaefb2e27f5e75ac8f9784b494363.pdf
《Training Neural Machines with Partial Traces》:
关键词：Neural Abstract Machines Neural Turing Machines Neural Random Access Machines Program Synthesis Program Induction
下载地址：https://openreview.net/pdf/37c886cc0ba98deadbca18fa7315b0d28acb9359.pdf
《Faster Reinforcement Learning with Expert State Sequences》:
关键词：Reinforcement Learning Imitation Learning
下载地址：https://openreview.net/pdf/6dc921376ff46c9bc2055b92f9b2d2581c1fbdd0.pdf
《Spatially Transformed Adversarial Examples》:
关键词：adversarial examples spatial transformation
下载地址：https://openreview.net/pdf/be9e53c0a815f5a53679d886757da35a237d8e43.pdf
《Decision-Based Adversarial Attacks: reliable attacks against Black-Box Machine Learning Models》:
关键词：adversarial attacks adversarial examples adversarials
下载地址：https://openreview.net/pdf/be866f82b1ca2c8155aae2c1f1db9998cf7745ef.pdf
《Learning Priors for Adversarial Autoencoders》:
关键词：deep learning computer vision generative adversarial networks
下载地址：https://openreview.net/pdf/b65574c440d08f49236650a0ff404fad91f0d3bf.pdf
《The Information-Autoencoding Family: A Lagrangian Perspective on Latent Variable Generative Modeling》:
关键词：Generative Models Variational Autoencoder Generative Adversarial Network
下载地址：https://openreview.net/pdf/27ef4955b028c7a77d5b76170b017d35d32e5569.pdf
《Variance Regularizing Adversarial Learning》:
关键词：Generative Adversarial Network Integral Probability Metric Meta-Adversarial Learning
下载地址：https://openreview.net/pdf/40314ec625815016cf5f00ff379932bcc1815bb8.pdf
《Communication Algorithms via Deep Learning》:
关键词：coding theory recurrent neural network communication
下载地址：https://openreview.net/pdf/3aea5e46f4814781d3d3edbe98155d0a57723dff.pdf
《Gated ConvNets for Letter-Based ASR》:
关键词：automatic speech recognition letter-based acoustic model gated convnets
下载地址：https://openreview.net/pdf/fd0e5b0202837238db3b2fbaf15b614696ade0a1.pdf
《On Characterizing the Capacity of Neural Networks Using Algebraic Topology》:
关键词：deep learning theory architecture selection algebraic topology
下载地址：https://openreview.net/pdf/c76ab60451a21aa9280286d17cb201160fb5cbf4.pdf
《Towards Image Understanding from Deep Compression Without Decoding》:
下载地址：https://openreview.net/pdf/ab1882c18d655b75d232dd208c49aa09b3ccb4ac.pdf
《TOWARDS SAFE DEEP LEARNING: UNSUPERVISED DEFENSE AGAINST GENERIC ADVERSARIAL ATTACKS》:
关键词：UNSUPERVISED DEFENSE ADVERSARIAL ATTACKS DEEP LEARNING
下载地址：https://openreview.net/pdf/a5c6585534d2cc7f52f0ead4671b373fbd8c54cc.pdf
《DEMYSTIFYING WIDE NONLINEAR AUTO-ENCODERS: FAST SGD CONVERGENCE TOWARDS SPARSE REPRESENTATION FROM RANDOM INITIALIZATION》:
关键词：stochastic gradient descent autoencoders nonconvex optimization representation learning theory
下载地址：https://openreview.net/pdf/4b61e03c354a1c7f2ab7ec1f071d53c9ba111a40.pdf
《Unsupervised Machine Translation Using Monolingual Corpora Only》:
关键词：unsupervised machine translation adversarial
下载地址：https://openreview.net/pdf/1d802bb8648d72779a723e7b582de5cb6b5331df.pdf
《Boosting the Actor with Dual Critic》:
关键词：reinforcement learning actor-critic algorithm Lagrangian duality
下载地址：https://openreview.net/pdf/53558e0d7d47dc1c8f1cea8f558c3f65315afc8a.pdf
《Visualizing the Loss Landscape of Neural Nets》:
下载地址：https://openreview.net/pdf/2f8378bb7e7e0c243165ab91d87eae4eaea8698f.pdf
《Transfer Learning on Manifolds via Learned Transport Operators》:
关键词：manifold learning transfer learning
下载地址：https://openreview.net/pdf/af26e6ce03b387cd8eba7856931f559cca818ad2.pdf
《Predictions》:
下载地址：https://openreview.net/pdf/e201dfe8ebc97fcb5797f0180798e1661464466f.pdf
《Jiffy: A Convolutional Approach to Learning Time Series Similarity》:
关键词：Time Series Time Series Classification
下载地址：https://openreview.net/pdf/54a039ab00baa6b5f3d6294973485db8468899db.pdf
《Parametrizing filters of a CNN with a GAN》:
关键词：invariance cnn gan infogan transformation
下载地址：https://openreview.net/pdf/1417b939577e252c1f8c08b510a5c3268c28f4d1.pdf
《Improving diversity in Generative adversarial networks by encouraging discriminator representation entropy》:
下载地址：https://openreview.net/pdf/ff9140585e648167a7db15960dfff8b4cc505bff.pdf
《Distributional Policy Gradients》:
关键词：policy gradient continuous control actor critic reinforcement learning
下载地址：https://openreview.net/pdf/e3cba47803f92fe6868f4b2605bcdf1b60118f21.pdf
《Quadrature-based features for kernel approximation》:
关键词：kernel methods low-rank approximation quadrature rules random features
下载地址：https://openreview.net/pdf/43a72416f1d6f0831a45ae4c454ee293eceb5757.pdf
《AUTOMATA GUIDED HIERARCHICAL REINFORCE- MENT LEARNING FOR ZERO-SHOT SKILL COMPOSI- TION》:
关键词：Hierarchical Reinforcement learning temporal logic skill composition
下载地址：https://openreview.net/pdf/a97cdce29399daece7fc8fa7a274c8e4613e15fb.pdf
《ShakeDrop regularization》:
下载地址：https://openreview.net/pdf/04315033857a1806baf59b2871c7375c263aa65d.pdf
《Clustering with Deep Learning: Taxonomy and New Methods》:
关键词：clustering deep learning neural networks
下载地址：https://openreview.net/pdf/e03d46184bc51997e6756874bdcba2cedcfa8606.pdf
《Bit-Regularized Optimization of Neural Nets》:
下载地址：https://openreview.net/pdf/535907a85d12776d0535edc83adf5509084b2f86.pdf
《Key Protected Classification for GAN Attack Resilient Collaborative Learning》:
关键词：privacy preserving deep learning collaborative learning adversarial attack
下载地址：https://openreview.net/pdf/bafff473e16c4115d8fcd80cb0d5047f430aa851.pdf
《Cluster-based Warm-Start Nets》:
关键词：hierarchical labels weak labels pairwise constraints clustering classification
下载地址：https://openreview.net/pdf/1b36b2c77f1d79b4e8e8483cd5f9cdced3d6444c.pdf
《Learning to Write by Learning the Objective》:
关键词：natural language generation
下载地址：https://openreview.net/pdf/944843774a69d1f438a216e893ed052c9298b83b.pdf
《A DIRT-T Approach to Unsupervised Domain Adaptation》:
关键词：domain adaptation unsupervised learning semi-supervised learning
下载地址：https://openreview.net/pdf/5d214b5eddd3ea75309ac7e2b2c203a4d01fd636.pdf
《REGULARIZATION NEURAL NETWORKS VIA CONSTRAINED VIRTUAL MOVEMENT FILED》:
下载地址：https://openreview.net/pdf/e4af05ee205636e7cfcbf612220dafde4fe1170e.pdf
《TOWARDS A GENERALIZATION THEORY AND TEST FOR GENERATIVE ADVERSARIAL NETWORKS》:
关键词：generative adversarial networks Wasserstein GAN generalization theory
下载地址：https://openreview.net/pdf/cb4e04739bca241d27f32eedd407926036aeaec9.pdf
《Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs》:
关键词：GAN medical records time series generation privacy
下载地址：https://openreview.net/pdf/a6406e34205961f2f85966b096fada789313aa89.pdf
《Forward Modeling for Partial Observation Strategy Games - A StarCraft Defogger》:
关键词：forward modeling partially observable deep learning strategy game real-time strategy
下载地址：https://openreview.net/pdf/aae1d23f51026ecf95f86352aee60f28c30928b5.pdf
《ElimiNet: A Model for Eliminating Options for Reading Comprehension with Multiple Choice Questions》:
关键词：Reading Comprehension Answering Multiple Choice Questions
下载地址：https://openreview.net/pdf/ec9693345d75f670ab35c40974afc583f0f4d12f.pdf
《Self-Supervised Learning of Object Motion Through Adversarial Video Prediction》:
关键词：adversarial video prediction flow
下载地址：https://openreview.net/pdf/f12ed6b3f7d214e767e4bb91ab2987bdf3569a6c.pdf
《LEARNING TO SHARE: SIMULTANEOUS PARAMETER TYING AND SPARSIFICATION IN DEEP LEARNING》:
关键词：Compressing neural network simultaneously parameter tying and sparsification group ordered l1 regularization
下载地址：https://openreview.net/pdf/de66b05d0b99c974929def1ab7fe115679c6582f.pdf
《Reinforcement Learning on Web Interfaces using Workflow-Guided Exploration》:
关键词：reinforcement learning sparse rewards web exploration
下载地址：https://openreview.net/pdf/b1663898b2b7a6a4f5e649a9db8949078b349f2d.pdf
《Building Generalizable Agents with a Realistic and Rich 3D Environment》:
关键词：reinforcement learning navigation generalization 3D scenes
下载地址：https://openreview.net/pdf/7c8c6d6e8eea4f04fe607eb0dabfdbba1110316b.pdf
《LEARNING A GENERATIVE MODEL FOR VALIDITY IN COMPLEX DISCRETE STRUCTURES》:
关键词：Active learning Reinforcement learning Molecules
下载地址：https://openreview.net/pdf/1f642e73faf93087c5ce2057b10fd02ed2017442.pdf
《Multimodal Sentiment Analysis To Explore the Structure of Emotions》:
下载地址：https://openreview.net/pdf/c00807a22d7e64009951a086257ff4f8b0f763be.pdf
《Parameterized Hierarchical Procedures for Neural Programming》:
关键词：Neural programming Hierarchical Control
下载地址：https://openreview.net/pdf/4007b99586e4f95de8cd1a8d192f374e1559aa2b.pdf
《Auto-Encoding Sequential Monte Carlo》:
关键词：Variational Autoencoders Inference amortization Model learning Sequential Monte Carlo ELBOs
下载地址：https://openreview.net/pdf/1336e9629689e1bc9c007ab1e51fe412ca3581c8.pdf
《EXPLORING NEURAL ARCHITECTURE SEARCH FOR LANGUAGE TASKS》:
关键词：Neural architecture search language tasks neural machine translation reading comprehension SQuAD
下载地址：https://openreview.net/pdf/14cabdce2b24dabb021947efc915897b476af354.pdf
《MACHINE VS MACHINE: DEFENDING CLASSIFIERS AGAINST LEARNING-BASED ADVERSARIAL ATTACKS》:
下载地址：https://openreview.net/pdf/be1767c6461d00041c1083a39b9e3d4095956260.pdf
《A dynamic game approach to training robust deep policies》:
关键词：game-theory reinforcement-learning guided-policy-search dynamic-programming
下载地址：https://openreview.net/pdf/fc0d41d3e64c58ed947de9a5b43e2ec219a857a9.pdf
《Learning Sparse Structured Ensembles with SG-MCMC and Network Pruning》:
关键词：ensemble learning SG-MCMC group sparse prior network pruning
下载地址：https://openreview.net/pdf/d2c875aa64cf80b7259ff1219d9b6450536bbcdb.pdf
《Syntax-Directed Variational Autoencoder for Structured Data》:
关键词：generative model for structured data syntax-directed generation molecule and program optimization variational autoencoder
下载地址：https://openreview.net/pdf/bcb84d5d815b673a07f8784ab66d446ba4069f2a.pdf
《Model Specialization for Inference Via End-to-End Distillation, Pruning, and Cascades》:
下载地址：https://openreview.net/pdf/2b76ff3ae94fd9426c6f888a25a981da432a56ba.pdf
《TRUNCATED HORIZON POLICY SEARCH: DEEP COMBINATION OF REINFORCEMENT AND IMITATION》:
关键词：Imitation Learning Reinforcement Learning
下载地址：https://openreview.net/pdf/cf252bd5d6a654b7acfc39280eb7771c52fdf486.pdf
《Adversarial Examples for Natural Language Classification Problems》:
下载地址：https://openreview.net/pdf/af7bad9ab688e80ec5aa6670c811c585d26d24dd.pdf
《Learn to Pay Attention》:
关键词：deep learning attention-aware representations image classification weakly supervised segmentation domain shift classifier generalisation robustness to adversarial attack
下载地址：https://openreview.net/pdf/83c8388e456aca6cdead7f7e65849b828ea46022.pdf
《Generalization of Learning using Reservoir Computing》:
关键词：Generalization Reservoir Computing dynamical system Siamese Neural Network image classification similarity dimensionality reduction
下载地址：https://openreview.net/pdf/8bf870b2b07b44ac87e15008f27d94ed52297932.pdf
《Flipout: Efficient Pseudo-Independent Weight Perturbations on Mini-Batches》:
关键词：weight perturbation reparameterization gradient gradient variance reduction evolution strategies
下载地址：https://openreview.net/pdf/7301873434d1a5bb65f679d06881c0fe9917cb6d.pdf
《On the difference between building and extracting patterns: a causal analysis of deep generative models.》:
关键词：GAN VAE causality
下载地址：https://openreview.net/pdf/cb1b4a4f05d77804dc36f7b279fe367df0aca8e6.pdf
《Jointly Learning to Construct and Control Agents using Deep Reinforcement Learning》:
关键词：robot locomotion reinforcement learning policy gradients physical design deep learning
下载地址：https://openreview.net/pdf/61209d2fa0c99fe17b00d5e5c2decb02599cf2f4.pdf
《Tandem Blocks in Deep Convolutional Neural Networks》:
关键词：resnet residual shortcut convolutional linear skip highway
下载地址：https://openreview.net/pdf/90d941323467bef7e116ae88765b00da6522e10a.pdf
《Analyzing GANs with Generative Scattering Networks》:
关键词：Unsupervised Learning Inverse Problems Convolutional Networks Generative Models Scattering Transform
下载地址：https://openreview.net/pdf/aef590d297e1be54323030268ce041a461ded8b7.pdf
《Lifelong Learning with Dynamically Expandable Networks》:
关键词：Transfer learning Lifelong learning Selective retraining Dynamic network expansion
下载地址：https://openreview.net/pdf/63d0ae2b3f179b39aa548d821c9922de16a3b081.pdf
《Bounding and Counting Linear Regions of Deep Neural Networks》:
关键词：rectifier networks maxout networks piecewise linear functions linear regions mixed-integer programming
下载地址：https://openreview.net/pdf/d32bf700a1a75b548333244a3f440231dc35435b.pdf
《Deep Boosting of Diverse Experts》:
关键词：boosting learning deep learning neural network
下载地址：https://openreview.net/pdf/1bf8847d8795430bde1519605578e2fad6d35ee7.pdf
《LSH-SAMPLING BREAKS THE COMPUTATIONAL CHICKEN-AND-EGG LOOP IN ADAPTIVE STOCHASTIC GRADIENT ESTIMATION》:
关键词：Stochastic Gradient Descent Optimization Sampling Estimation
下载地址：https://openreview.net/pdf/4a6a47622cca390bf07da0e029487cfe84e80f5d.pdf
《Network of Graph Convolutional Networks \ Trained on Random Walks》:
关键词：Graph Convolution Deep Learning Network of Networks
下载地址：https://openreview.net/pdf/183183d6f43081020398c458aef6f341a2ee9619.pdf
《Neural Tree Transducers for Tree to Tree Learning》:
关键词：deep learning tree transduction
下载地址：https://openreview.net/pdf/297b7ddaf5c031893d7c9f0e6587fddd20150379.pdf
《Simulated+Unsupervised Learning With Adaptive Data Generation and Bidirectional Mappings》:
下载地址：https://openreview.net/pdf/20c79150ce7adc491a548d89f07d22917313b079.pdf
《Gaussian Process Neurons》:
关键词：gaussian process neuron activation function stochastic transfer function learning variational bayes probabilistic
下载地址：https://openreview.net/pdf/ade247befee7fc59f5293b8e372be246ef9e3cc3.pdf
《Deep Hyperspherical Defense against Adversarial Perturbations》:
下载地址：https://openreview.net/pdf/1e36d567ac45167e2997f76d2c4f23f631f05d08.pdf
《Curiosity-driven Exploration by Bootstrapping Features》:
关键词：exploration intrinsic motivation reinforcement learning
下载地址：https://openreview.net/pdf/7aab3713419f7b614e7aabc00779f7d5502f69f5.pdf
《FastNorm: Improving Numerical Stability of Deep Network Training with Efficient Normalization》:
关键词：Neural networks Training Convergence
下载地址：https://openreview.net/pdf/f99a16bc7c712859a8c0167f7615f33c18c9d218.pdf
《DETECTING ADVERSARIAL PERTURBATIONSWITH SALIENCY》:
关键词：DETECTING ADVERSARIAL PERTURBATIONSWITH SALIENCY
下载地址：https://openreview.net/pdf/973e7c7b32e0a102724f8388caca59cfe8b0e9ec.pdf
《Fast and Accurate Text Classification: Skimming, Rereading and Early Stopping》:
关键词：Topic Classification Sentiment Analysis Natural Language Processing
下载地址：https://openreview.net/pdf/748926bc02b774ab9382a6589bd0940a11059f4b.pdf
《The power of deeper networks for expressing natural functions》:
关键词：expressivity of neural networks depth of neural networks universal approximators function approximation deep learning
下载地址：https://openreview.net/pdf/f1e0407ce38819eb7ab40b4349184e630227370f.pdf
《Improving generalization with Wasserstein regularization》:
关键词：natural gradient generalization optimization
下载地址：https://openreview.net/pdf/8136b4f93b54ad266acbedc42f1ef4505bc0d114.pdf
《DeepArchitect: Automatically Designing and Training Deep Architectures》:
关键词：architecture search deep learning hyperparameter tuning
下载地址：https://openreview.net/pdf/1b4b11e6581fddc7fc91f511087335cef3ad7fb1.pdf
《Deep Learning for Physical Processes: Incorporating Prior Scientific Knowledge》:
关键词：deep learning physical processes forecasting spatio-temporal
下载地址：https://openreview.net/pdf/02a9a233a7aa971149a7f6ea697a4a152fa3689a.pdf
《Few-shot learning with simplex》:
关键词：one-shot learning few-shot learning deep learning simplex
下载地址：https://openreview.net/pdf/fddb579342c402e865d07408e33a91d47f614525.pdf
《Quantitatively Evaluating GANs With Divergences Proposed for Training》:
关键词：Generative adversarial networks
下载地址：https://openreview.net/pdf/578a04b8857b7d7cc66ce81581dc871a67f6e3e4.pdf
《Faster Distributed Synchronous SGD with Weak Synchronization》:
关键词：distributed deep learning straggler
下载地址：https://openreview.net/pdf/1c1ff517c33f6e17855efaec19617cddbfe2b617.pdf
《WHAI: Weibull Hybrid Autoencoding Inference for Deep Topic Modeling》:
下载地址：https://openreview.net/pdf/2ef4a68e4966240f14358e2b7f060729de25a5c6.pdf
《Latent forward model for Real-time Strategy game planning with incomplete information》:
关键词：Real time strategy latent space forward model monte carlo tree search reinforcement learning planning
下载地址：https://openreview.net/pdf/1ab4ab0a8c425e3c3dfbab7708bec527a4bc0029.pdf
《Learning to Treat Sepsis with Multi-Output Gaussian Process Deep Recurrent Q-Networks》:
关键词：Healthcare Gaussian Process Deep Reinforcement Learning
下载地址：https://openreview.net/pdf/b38ae962f024e0b70a031a91f61e56b9b3b37a00.pdf
《Few-Shot Learning with Graph Neural Networks》:
下载地址：https://openreview.net/pdf/a607d91d8b41622cf3c1e29e1de2eb1bb27222fa.pdf
《Automatic Parameter Tying in Neural Networks》:
关键词：neural network quantization compression
下载地址：https://openreview.net/pdf/05ada7c9366bbaa485bc160372d791a3796faecd.pdf
《Divide and Conquer Networks》:
关键词：Neural Networks Combinatorial Optimization Algorithms
下载地址：https://openreview.net/pdf/69e2371a976edc0d109b604168709fdda5c6823b.pdf
《A comparison of second-order methods for deep convolutional neural networks》:
下载地址：https://openreview.net/pdf/88333cf43fb99cea205b2a0b0f344e4d66abd756.pdf
《Lifelong Word Embedding via Meta-Learning》:
关键词：Lifelong learning meta learning word embedding
下载地址：https://openreview.net/pdf/8869339741d516a02d6aa995cd6911ea6f599a13.pdf
《Extending the Framework of Equilibrium Propagation to General Dynamics》:
关键词：Deep Learning Backpropagation Fixed Point Recurrent Neural Network Biologically Plausible Learning Feedback Alignment Dynamical System Gradient-Free Optimization
下载地址：https://openreview.net/pdf/16a90cccd4649a3c968d342f202e5f21be2ea2be.pdf
《Synthesizing Robust Adversarial Examples》:
关键词：adversarial examples
下载地址：https://openreview.net/pdf/5d7a5a26f87b447e17ef5c08b6393a6d9ff878a6.pdf
《Understanding GANs: the LQG Setting》:
关键词：Generative Adversarial Networks Wasserstein Generalization PCA
下载地址：https://openreview.net/pdf/5fd49847c822b7cc58efd57865bf98665a3abd3b.pdf
《Bayesian Embeddings for Long-Tailed Datasets》:
关键词：Long-tail datasets Imbalanced datasets
下载地址：https://openreview.net/pdf/eb02d59c3547196c9f6bdfcdf8382978fd8dde70.pdf
《Distributed Restarting NewtonCG Method for Large-Scale Empirical Risk Minimization》:
下载地址：https://openreview.net/pdf/38836ab2cd82ed11d6d1b6aa9fdd3b0e02435ce4.pdf
《Generative Discovery of Relational Medical Entity Pairs》:
关键词：Knowledge Discovery Generative Modeling Medical Entity Pair
下载地址：https://openreview.net/pdf/1b18496cb3dd605b82c84d2db7d3ac4387f7c56d.pdf
《Representing Entropy : A short proof of the equivalence between soft Q-learning and policy gradients》:
关键词：soft Q-learning policy gradients entropy Legendre transformation duality convex analysis Donsker-Varadhan
下载地址：https://openreview.net/pdf/d8235f3f5648e3f3f3f2a8ea7fd28d1e533f1d1e.pdf
《A Semantic Loss Function for Deep Learning with Symbolic Knowledge》:
关键词：deep learning symbolic knowledge semi-supervised learning constraints
下载地址：https://openreview.net/pdf/432d927483521f91358d6dc63833a1973ab1ef5c.pdf
《Hierarchical Representations for Efficient Architecture Search》:
关键词：deep learning architecture search
下载地址：https://openreview.net/pdf/8d170a17a92217a270f4cdfbe1c84aba3c812530.pdf
《On the insufficiency of existing momentum schemes for Stochastic Optimization》:
关键词：Stochastic Gradient Descent Deep Learning Momentum Acceleration Heavy Ball Nesterov Acceleration Stochastic Optimization SGD
下载地址：https://openreview.net/pdf/2a3ed17757044e13a22615ff1fa39608b8ac0d46.pdf
《Incremental Learning in Deep Convolutional Neural Networks Using Partial Network Sharing》:
关键词：Deep learning Incremental learning energy-efficient learning supervised learning
下载地址：https://openreview.net/pdf/8daed5bfae4170779872ce77a2b14b5d81a84027.pdf
《Attacking Binarized Neural Networks》:
关键词：adversarial examples adversarial attacks binary binarized neural networks
下载地址：https://openreview.net/pdf/905e88c170ffde8b30ec716ef22f4b949d5eaa79.pdf
《Attention-based Graph Neural Network for Semi-supervised Learning》:
关键词：Graph Neural Network Attention Semi-supervised
下载地址：https://openreview.net/pdf/2171447de61701922ab934d1bcfac5f0a1274455.pdf
《Variational image compression with a scale hyperprior》:
下载地址：https://openreview.net/pdf/fe1790de2c4943e3fff4ca04bf7dec999b5c8ecf.pdf
《Beyond Shared Hierarchies: Deep Multitask Learning through Soft Layer Ordering》:
关键词：multitask learning deep learning modularity
下载地址：https://openreview.net/pdf/5a794213cb2b9a374d1ef348a9091715c49648f5.pdf
《Forced Apart: Discovering Disentangled Representations Without Exhaustive Labels》:
关键词：learning representation clustering loss
下载地址：https://openreview.net/pdf/1127adbc6a9f6d01b20b784518adc30170f05d58.pdf
《Weightless: Lossy Weight Encoding For Deep Neural Network Compression》:
关键词：Deep Neural Network Compression Sparsity
下载地址：https://openreview.net/pdf/853cfd55f56c721fc8f6c689cb549bdf6c9c9809.pdf
《Sparse Deep Scattering Croisé Network》:
关键词：Deep Scattering Network Continuous Wavelet Thresholding Sparse Activations Time-frequency represenation Multi-Family Wavelets Convolutional Network Bird Detection
下载地址：https://openreview.net/pdf/71574323fc19eb7fda5722193c5c7e18d2b94fc2.pdf
《Go for a Walk and Arrive at the Answer: Reasoning Over Paths in Knowledge Bases using Reinforcement Learning》:
关键词：Knowledge Graphs Reinforcement Learning Query Answering
下载地址：https://openreview.net/pdf/f4a51e16d80de24a14656f9a0e31df63a6bccaf6.pdf
《Prototype Matching Networks for Large-Scale Multi-label Genomic Sequence Classification》:
关键词：bioinformatics multi-label classification matching networks prototypes memory networks attention
下载地址：https://openreview.net/pdf/ae4cfb2d369c4359baf433a390382fa790b5bb48.pdf
《LEAP: Learning Embeddings for Adaptive Pace》:
关键词：deep metric learning self-paced learning representation learning cnn
下载地址：https://openreview.net/pdf/f4e80762d5c1fec1b5d8b82c8ea95b76476f740c.pdf
《GraphGAN: Generating Graphs via Random Walks》:
关键词：GAN graphs random walks implicit generative models
下载地址：https://openreview.net/pdf/97f95191af801049263661e8d0c4c1dd560c5578.pdf
《Zero-Shot Visual Imitation》:
关键词：imitation zero shot self-supervised robotics skills navigation manipulation
下载地址：https://openreview.net/pdf/033006fc0917363d809a60477a753aecc800ddf0.pdf
《Sparse Attentive Backtracking: Long-Range Credit Assignment in Recurrent Networks》:
关键词：recurrent neural networks long-term dependencies back-propagation through time truncated back-propagation biological inspiration self-attention
下载地址：https://openreview.net/pdf/17aa0f1cb740cd2e0bc8da2c8f614a7cba5a6678.pdf
《Mixed Precision Training of Convolutional Neural Networks using Integer Operations》:
关键词：deep learning training reduced precision imagenet dynamic fixed point
下载地址：https://openreview.net/pdf/c412c6aeeebf58b37da02b47cf075aa2f6f71ff2.pdf
《Imitation Learning from Visual Data with Multiple Intentions》:
关键词：multi-modal imitation learning deep learning generative models stochastic neural networks
下载地址：https://openreview.net/pdf/74f194700bba77103e3056915ba765c2139e8a33.pdf
《Demystifying MMD GANs》:
关键词：gans mmd ipms wgan gradient penalty unbiased gradients
下载地址：https://openreview.net/pdf/37e2db86ed222ed46afeaeee8ee9e22ba459d7eb.pdf
《Reward Estimation via State Prediction》:
关键词：reinforcement learning inverse reinforcement learning imitation learning
下载地址：https://openreview.net/pdf/81ead848e9128bdba0bf62ac219b44ae6759df39.pdf
《Predict Responsibly: Increasing Fairness by Learning to Defer》:
关键词：Fairness IDK Calibration Automated decision-making Transparency Accountability
下载地址：https://openreview.net/pdf/19fbc75438c53be785862db3a715309f0f668657.pdf
《A framework for the quantitative evaluation of disentangled representations》:
下载地址：https://openreview.net/pdf/b2840e4750cf351b8396c960770cbe90d0fda1b3.pdf
《Learning Document Embeddings With CNNs》:
关键词：unsupervised embedding convolutional neural network
下载地址：https://openreview.net/pdf/7dde4526398d464473538e0cfaaf34af06538728.pdf
因篇幅限制，请到专知网站查看获者；原始链接读取。
专知网站查看ICLR2018提交论文：
请登录www.zhuanzhi.ai或者点击阅读原文，顶端搜索“ICLR” 主题，取查看获得专知全网收录资源, 涵盖动态资讯、干货等面经！如下图所示~
此外，请登录www.zhuanzhi.ai或者点击阅读原文，顶端搜索“深度学习” 主题，取查看获得专知全网收录资源, 涵盖动态资讯、干货等面经！如下图所示~
获取ICLR2018提交论文 pdf文件 以及深度学习资料大全：
请关注专知公众号（扫一扫最下面专知二维码，或者点击上方蓝色专知），
后台回复“LCLR” 就可以获取 LCLR2018匿名提交论文 的pd下载链接~~；
后台回复“MLDL” 就可以获取 机器学习&深度学习知识资料大全集的pdf第一篇下载链接~~；
后台回复“MLDL2” 就可以获取 机器学习&深度学习知识资料大全集的pdf第二篇下载链接~~
更多专知荟萃知识资料全集获取，请查看：
【干货荟萃】机器学习&深度学习知识资料大全集(一)（论文/教程/代码/书籍/数据/课程等）
【GAN货】生成对抗网络知识资料全集(论文/代码/教程/视频/文章等)
【干货】Google GAN之父Ian Goodfellow ICCV2017演讲：解读生成对抗网络的原理与应用
【AlphaGoZero核心技术】深度强化学习知识资料全集(论文/代码/教程/视频/文章等)
欢迎转发到你的微信群和朋友圈，分享专业AI知识！
获取更多关于机器学习以及人工智能知识资料，请访问www.zhuanzhi.ai, 或者点击阅读原文，即可得到！
-END-
欢迎使用专知
专知，一个新的认知方式！目前聚焦在人工智能领域为AI从业者提供专业可信的知识分发服务, 包括主题定制、主题链路、搜索发现等服务，帮你又好又快找到所需知识。
使用方法>>访问www.zhuanzhi.ai, 或点击文章下方“阅读原文”即可访问专知
中国科学院自动化研究所专知团队
@2017 专知
专 · 知
关注我们的公众号，获取最新关于专知以及人工智能的资讯、技术、算法、深度干货等内容。扫一扫下方关注我们的微信公众号。
点击“阅读原文”，使用专知！