ICLR 2017

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

iclr2017:conference_posters [2017/03/28 08:20]
rnogueira
iclr2017:conference_posters [2017/04/23 09:26] (current)
hugo
Line 3: Line 3:
 Below are the Conference Track papers presented at each of the poster sessions (on Monday, Tuesday or Wednesday, in the morning or evening). To find a paper, look for the poster with the corresponding number in the area dedicated to the Conference Track. Below are the Conference Track papers presented at each of the poster sessions (on Monday, Tuesday or Wednesday, in the morning or evening). To find a paper, look for the poster with the corresponding number in the area dedicated to the Conference Track.
  
 +======Note to the Presenters=======
 +Each poster panel is 2 meters large and 1 meter tall.\\
 +If needed, tape will be provided to fix your poster.
 +
 +
 +<​html><​div id='​monday_morning'></​div></​html>​
 ====Monday Morning (April 24th, 10:30am to 12:​30pm)==== ====Monday Morning (April 24th, 10:30am to 12:​30pm)====
-  - Making Neural Programming Architectures Generalize via Recursion +C1: Making Neural Programming Architectures Generalize via Recursion\\ 
-  ​- ​Learning Graphical State Transitions +C2: Learning Graphical State Transitions\\ 
-  ​- ​Distributed Second-Order Optimization using Kronecker-Factored Approximations +C3: Distributed Second-Order Optimization using Kronecker-Factored Approximations\\ 
-  ​- ​Normalizing the Normalizers:​ Comparing and Extending Network Normalization Schemes +C4: Normalizing the Normalizers:​ Comparing and Extending Network Normalization Schemes\\ 
-  ​- ​Neural Program Lattices +C5: Neural Program Lattices\\ 
-  ​- ​Diet Networks: Thin Parameters for Fat Genomics +C6: Diet Networks: Thin Parameters for Fat Genomics\\ 
-  ​- ​Unsupervised Cross-Domain Image Generation +C7: Unsupervised Cross-Domain Image Generation\\ 
-  ​- ​Towards Principled Methods for Training Generative Adversarial Networks +C8: Towards Principled Methods for Training Generative Adversarial Networks\\ 
-  ​- ​Recurrent Mixture Density Network for Spatiotemporal Visual Attention +C9: Recurrent Mixture Density Network for Spatiotemporal Visual Attention\\ 
-  ​- ​Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfer +C10: Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfer\\ 
-  ​- ​Pruning Filters for Efficient ConvNets +C11: Pruning Filters for Efficient ConvNets\\ 
-  Optimization as a Model for Few-Shot ​Learning +C12: Stick-Breaking Variational Autoencoders\\ 
-  - Understanding deep learning requires rethinking generalization +C13: Identity Matters in Deep Learning\\ 
-  - On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima +C14: On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima\\ 
-  ​- ​Recurrent Hidden Semi-Markov Model +C15: Recurrent Hidden Semi-Markov Model\\ 
-  ​- ​Nonparametric Neural Networks +C16: Nonparametric Neural Networks\\ 
-  ​- ​Learning to Generate Samples from Noise through Infusion Training +C17: Learning to Generate Samples from Noise through Infusion Training\\ 
-  ​- ​An Information-Theoretic Framework for Fast and Robust Unsupervised Learning via Neural Population Infomax +C18: An Information-Theoretic Framework for Fast and Robust Unsupervised Learning via Neural Population Infomax\\ 
-  ​- ​Highway and Residual Networks learn Unrolled Iterative Estimation +C19: Highway and Residual Networks learn Unrolled Iterative Estimation\\ 
-  ​- ​Soft Weight-Sharing for Neural Network Compression +C20: Soft Weight-Sharing for Neural Network Compression\\ 
-  ​- ​Snapshot Ensembles: Train 1, Get M for Free +C21: Snapshot Ensembles: Train 1, Get M for Free\\ 
-  ​- ​Towards a Neural Statistician +C22: Towards a Neural Statistician\\ 
-  ​- ​Learning Curve Prediction with Bayesian Neural Networks +C23: Learning Curve Prediction with Bayesian Neural Networks\\ 
-  ​- ​Learning End-to-End Goal-Oriented Dialog +C24: Learning End-to-End Goal-Oriented Dialog\\ 
-  ​- ​Multi-Agent Cooperation and the Emergence of (Natural) Language +C25: Multi-Agent Cooperation and the Emergence of (Natural) Language\\ 
-  ​- ​Efficient Vector Representation for Documents through Corruption +C26: Efficient Vector Representation for Documents through Corruption\\ 
-  ​- ​Improving Neural Language Models with a Continuous Cache +C27: Improving Neural Language Models with a Continuous Cache\\ 
-  ​- ​Program Synthesis for Character Level Language Modeling +C28: Program Synthesis for Character Level Language Modeling\\ 
-  ​- ​Tracking the World State with Recurrent Entity Networks +C29: Tracking the World State with Recurrent Entity Networks\\ 
-  ​- ​Reinforcement Learning with Unsupervised Auxiliary Tasks +C30: Reinforcement Learning with Unsupervised Auxiliary Tasks\\ 
-  ​- ​Neural Architecture Search with Reinforcement Learning +C31: Neural Architecture Search with Reinforcement Learning\\ 
-  ​- ​Sample Efficient Actor-Critic with Experience Replay +C32: Sample Efficient Actor-Critic with Experience Replay\\ 
-  ​- ​Learning to Act by Predicting the Future+C33: Learning to Act by Predicting the Future\\
  
 +<​html><​div id='​monday_afternoon'></​div></​html>​
 ====Monday Afternoon (April 24th, 4:30pm to 6:30pm)==== ====Monday Afternoon (April 24th, 4:30pm to 6:30pm)====
-  - Neuro-Symbolic Program Synthesis +C1: Neuro-Symbolic Program Synthesis\\ 
-  ​- ​Generative Models and Model Criticism via Optimized Maximum Mean Discrepancy +C2: Generative Models and Model Criticism via Optimized Maximum Mean Discrepancy\\ 
-  ​- ​Trained Ternary Quantization +C3: Trained Ternary Quantization\\ 
-  ​- ​DSD: Dense-Sparse-Dense Training for Deep Neural Networks +C4: DSD: Dense-Sparse-Dense Training for Deep Neural Networks\\ 
-  ​- ​A Compositional Object-Based Approach to Learning Physical Dynamics +C5: A Compositional Object-Based Approach to Learning Physical Dynamics\\ 
-  ​- ​Multilayer Recurrent Network Models of Primate Retinal Ganglion Cells +C6: Multilayer Recurrent Network Models of Primate Retinal Ganglion Cells\\ 
-  ​- ​Improving Generative Adversarial Networks with Denoising Feature Matching +C7: Improving Generative Adversarial Networks with Denoising Feature Matching\\ 
-  ​- ​Transfer of View-manifold Learning to Similarity Perception of Novel Objects +C8: Transfer of View-manifold Learning to Similarity Perception of Novel Objects\\ 
-  ​- ​What does it take to generate natural textures? +C9: What does it take to generate natural textures?\\ 
-  ​- ​Emergence of foveal image sampling from learning to attend in visual scenes +C10: Emergence of foveal image sampling from learning to attend in visual scenes\\ 
-  ​- ​PixelCNN++: A PixelCNN Implementation with Discretized Logistic Mixture Likelihood and Other Modifications +C11: PixelCNN++: A PixelCNN Implementation with Discretized Logistic Mixture Likelihood and Other Modifications\\ 
-  ​- ​Learning to Optimize +C12: Learning to Optimize\\ 
-  - Training Compressed Fully-Connected Networks with a Density-Diversity Penalty +C13: Do Deep Convolutional Nets Really Need to be Deep and Convolutional?​\\ 
-  ​- ​Optimal Binary Autoencoding with Pairwise Correlations +C14: Optimal Binary Autoencoding with Pairwise Correlations\\ 
-  ​- ​On the Quantitative Analysis of Decoder-Based Generative Models +C15: On the Quantitative Analysis of Decoder-Based Generative Models\\ 
-  - Learning to Remember Rare Events +C16: Adversarial machine learning at scale\\ 
-  ​- ​Transfer Learning for Sequence Tagging with Hierarchical Recurrent Networks +C17: Transfer Learning for Sequence Tagging with Hierarchical Recurrent Networks\\ 
-  ​- ​Capacity and Learnability in Recurrent Neural Networks +C18: Capacity and Learnability in Recurrent Neural Networks\\ 
-  ​- ​Deep Learning with Dynamic Computation Graphs +C19: Deep Learning with Dynamic Computation Graphs\\ 
-  ​- ​Exploring Sparsity in Recurrent Neural Networks +C20: Exploring Sparsity in Recurrent Neural Networks\\ 
-  ​- ​Structured Attention Networks +C21: Structured Attention Networks\\ 
-  - ZoneoutRegularizing RNNs by Randomly Preserving Hidden Activations +C22Learning to Repeat: Fine Grained Action Repetition for Deep Reinforcement Learning\\ 
-  ​- ​Variational Lossy Autoencoder +C23: Variational Lossy Autoencoder\\ 
-  ​- ​Learning to Query, Reason, and Answer Questions On Ambiguous Texts +C24: Learning to Query, Reason, and Answer Questions On Ambiguous Texts\\ 
-  ​- ​Deep Biaffine Attention for Neural Dependency Parsing +C25: Deep Biaffine Attention for Neural Dependency Parsing\\ 
-  ​- ​A Compare-Aggregate Model for Matching Text Sequences +C26: A Compare-Aggregate Model for Matching Text Sequences\\ 
-  ​- ​Data Noising as Smoothing in Neural Network Language Models +C27: Data Noising as Smoothing in Neural Network Language Models\\ 
-  ​- ​Neural Variational Inference For Topic Models +C28: Neural Variational Inference For Topic Models\\ 
-  - Words or Characters? Fine-grained Gating ​for Reading ​Comprehension +C29: Bidirectional Attention Flow for Machine ​Comprehension\\ 
-  ​- ​Q-Prop: Sample-Efficient Policy Gradient with An Off-Policy Critic +C30: Q-Prop: Sample-Efficient Policy Gradient with An Off-Policy Critic\\ 
-  ​- ​Stochastic Neural Networks for Hierarchical Reinforcement Learning +C31: Stochastic Neural Networks for Hierarchical Reinforcement Learning\\ 
-  ​- ​Learning Invariant Feature Spaces to Transfer Skills with Reinforcement Learning +C32: Learning Invariant Feature Spaces to Transfer Skills with Reinforcement Learning\\ 
-  ​- ​Third Person Imitation Learning+C33: Third Person Imitation Learning\\
  
 +<​html><​div id='​tuesday_morning'></​div></​html>​
 ====Tuesday Morning (April 25th, 10:30am to 12:​30pm)==== ====Tuesday Morning (April 25th, 10:30am to 12:​30pm)====
-  - DeepDSL: A Compilation-based Domain-Specific Language for Deep Learning +C1: DeepDSL: A Compilation-based Domain-Specific Language for Deep Learning\\ 
-  - SampleRNNAn Unconditional End-to-End Neural Audio Generation Model +C2A SELF-ATTENTIVE SENTENCE EMBEDDING\\ 
-  ​- ​Deep Probabilistic Programming +C3: Deep Probabilistic Programming\\ 
-  ​- ​Lie-Access Neural Turing Machines +C4: Lie-Access Neural Turing Machines\\ 
-  ​- ​Learning Features of Music From Scratch +C5: Learning Features of Music From Scratch\\ 
-  ​- ​Mode Regularized Generative Adversarial Networks +C6: Mode Regularized Generative Adversarial Networks\\ 
-  ​- ​End-to-end Optimized Image Compression +C7: End-to-end Optimized Image Compression\\ 
-  ​- ​Variational Recurrent Adversarial Deep Domain Adaptation +C8: Variational Recurrent Adversarial Deep Domain Adaptation\\ 
-  ​- ​Steerable CNNs +C9: Steerable CNNs\\ 
-  ​- ​Deep Predictive Coding Networks for Video Prediction and Unsupervised Learning +C10: Deep Predictive Coding Networks for Video Prediction and Unsupervised Learning\\ 
-  ​- ​PixelVAE: A Latent Variable Model for Natural Images +C11: PixelVAE: A Latent Variable Model for Natural Images\\ 
-  ​- ​A recurrent neural network without chaos +C12: A recurrent neural network without chaos\\ 
-  ​- ​Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer +C13: Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer\\ 
-  ​- ​Tree-structured decoding with doubly-recurrent neural networks +C14: Tree-structured decoding with doubly-recurrent neural networks\\ 
-  ​- ​Introspection:​Accelerating Neural Network Training By Learning Weight Evolution +C15: Introspection:​Accelerating Neural Network Training By Learning Weight Evolution\\ 
-  ​- ​Hyperband: Bandit-Based Configuration Evaluation for Hyperparameter Optimization +C16: Hyperband: Bandit-Based Configuration Evaluation for Hyperparameter Optimization\\ 
-  ​- ​Quasi-Recurrent Neural Networks +C17: Quasi-Recurrent Neural Networks\\ 
-  ​- ​Attend, Adapt and Transfer: Attentive Deep Architecture for Adaptive Transfer from multiple sources in the same domain +C18: Attend, Adapt and Transfer: Attentive Deep Architecture for Adaptive Transfer from multiple sources in the same domain\\ 
-  ​- ​A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks +C19: A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks\\ 
-  ​- ​Trusting SVM for Piecewise Linear CNNs +C20: Trusting SVM for Piecewise Linear CNNs\\ 
-  ​- ​Maximum Entropy Flow Networks +C21: Maximum Entropy Flow Networks\\ 
-  ​- ​The Concrete Distribution:​ A Continuous Relaxation of Discrete Random Variables +C22: The Concrete Distribution:​ A Continuous Relaxation of Discrete Random Variables\\ 
-  ​- ​Unrolled Generative Adversarial Networks +C23: Unrolled Generative Adversarial Networks\\ 
-  ​- ​A Simple but Tough-to-Beat Baseline for Sentence Embeddings +C24: A Simple but Tough-to-Beat Baseline for Sentence Embeddings\\ 
-  ​- ​Query-Reduction Networks for Question Answering +C25: Query-Reduction Networks for Question Answering\\ 
-  ​- ​Machine Comprehension Using Match-LSTM and Answer Pointer +C26: Machine Comprehension Using Match-LSTM and Answer Pointer\\ 
-  Bidirectional Attention Flow for Machine ​Comprehension +C27: Words or Characters? Fine-grained Gating ​for Reading ​Comprehension\\ 
-  ​- ​Dynamic Coattention Networks For Question Answering +C28: Dynamic Coattention Networks For Question Answering\\ 
-  ​- ​Multi-view Recurrent Neural Acoustic Word Embeddings +C29: Multi-view Recurrent Neural Acoustic Word Embeddings\\ 
-  ​- ​Episodic Exploration for Deep Deterministic Policies for StarCraft Micromanagement +C30: Episodic Exploration for Deep Deterministic Policies for StarCraft Micromanagement\\ 
-  ​- ​Training Agent for First-Person Shooter Game with Actor-Critic Curriculum Learning +C31: Training Agent for First-Person Shooter Game with Actor-Critic Curriculum Learning\\ 
-  ​- ​Generalizing Skills with Semi-Supervised Reinforcement Learning +C32: Generalizing Skills with Semi-Supervised Reinforcement Learning\\ 
-  ​- ​Improving Policy Gradient by Exploring Under-appreciated Rewards+C33: Improving Policy Gradient by Exploring Under-appreciated Rewards\\
  
-====Tuesday Afternoon (April 25th, 4:30pm to 6:30pm)==== +<​html><​div id='​tuesday_afternoon'></​div></​html>​ 
-  ​- ​Sigma Delta Quantized Networks +====Tuesday Afternoon (April 25th, 2:00pm to 4:00pm)==== 
-  ​- ​Paleo: A Performance Model for Deep Neural Networks +C1: Sigma Delta Quantized Networks\\ 
-  ​- ​DeepCoder: Learning to Write Programs +C2: Paleo: A Performance Model for Deep Neural Networks\\ 
-  ​- ​Topology and Geometry of Deep Rectified Network Optimization Landscapes +C3: DeepCoder: Learning to Write Programs\\ 
-  ​- ​Incremental Network Quantization:​ Towards Lossless CNNs with Low-precision Weights +C4: Topology and Geometry of Deep Rectified Network Optimization Landscapes\\ 
-  ​- ​Learning to Perform Physics Experiments via Deep Reinforcement Learning +C5: Incremental Network Quantization:​ Towards Lossless CNNs with Low-precision Weights\\ 
-  ​- ​Decomposing Motion and Content for Natural Video Sequence Prediction +C6: Learning to Perform Physics Experiments via Deep Reinforcement Learning\\ 
-  ​- ​Calibrating Energy-based Generative Adversarial Networks +C7: Decomposing Motion and Content for Natural Video Sequence Prediction\\ 
-  ​- ​Pruning Convolutional Neural Networks for Resource Efficient Inference +C8: Calibrating Energy-based Generative Adversarial Networks\\ 
-  ​- ​Incorporating long-range consistency in CNN-based texture generation +C9: Pruning Convolutional Neural Networks for Resource Efficient Inference\\ 
-  ​- ​Lossy Image Compression with Compressive Autoencoders +C10: Incorporating long-range consistency in CNN-based texture generation\\ 
-  ​- ​LR-GAN: Layered Recursive Generative Adversarial Networks for Image Generation +C11: Lossy Image Compression with Compressive Autoencoders\\ 
-  ​- ​Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data +C12: LR-GAN: Layered Recursive Generative Adversarial Networks for Image Generation\\ 
-  ​- ​Deep Variational Bayes Filters: Unsupervised Learning of State Space Models from Raw Data +C13: Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data\\ 
-  ​- ​Mollifying Networks +C14: Deep Variational Bayes Filters: Unsupervised Learning of State Space Models from Raw Data\\ 
-  ​- ​beta-VAE: Learning Basic Visual Concepts with a Constrained Variational Framework +C15: Mollifying Networks\\ 
-  ​- ​Categorical Reparameterization with Gumbel-Softmax +C16: beta-VAE: Learning Basic Visual Concepts with a Constrained Variational Framework\\ 
-  ​- ​Online Bayesian Transfer Learning for Sequential Data Modeling +C17: Categorical Reparameterization with Gumbel-Softmax\\ 
-  ​- ​Latent Sequence Decompositions +C18: Online Bayesian Transfer Learning for Sequential Data Modeling\\ 
-  ​- ​Density estimation using Real NVP +C19: Latent Sequence Decompositions\\ 
-  ​- ​Recurrent Batch Normalization +C20: Density estimation using Real NVP\\ 
-  ​- ​SGDR: Stochastic Gradient Descent with Restarts +C21: Recurrent Batch Normalization\\ 
-  ​- ​Variable Computation in Recurrent Neural Networks +C22: SGDR: Stochastic Gradient Descent with Restarts\\ 
-  ​- ​Deep Variational Information Bottleneck +C23: Variable Computation in Recurrent Neural Networks\\ 
-  A SELF-ATTENTIVE SENTENCE EMBEDDING +C24: Deep Variational Information Bottleneck\\ 
-  ​- ​TopicRNN: A Recurrent Neural Network with Long-Range Semantic Dependency +C25: SampleRNN: An Unconditional End-to-End Neural Audio Generation Model\\ 
-  ​- ​Frustratingly Short Attention Spans in Neural Language Modeling +C26: TopicRNN: A Recurrent Neural Network with Long-Range Semantic Dependency\\ 
-  ​- ​Offline Bilingual Word Vectors ​Without a Dictionary +C27: Frustratingly Short Attention Spans in Neural Language Modeling\\ 
-  ​- ​LEARNING A NATURAL LANGUAGE INTERFACE WITH NEURAL PROGRAMMER +C28: Offline Bilingual Word Vectors, Orthogonal Transformations and the Inverted Softmax\\ 
-  ​- ​Designing Neural Network Architectures using Reinforcement Learning +C29: LEARNING A NATURAL LANGUAGE INTERFACE WITH NEURAL PROGRAMMER\\ 
-  ​- ​Metacontrol for Adaptive Imagination-Based Optimization +C30: Designing Neural Network Architectures using Reinforcement Learning\\ 
-  ​- ​Recurrent Environment Simulators +C31: Metacontrol for Adaptive Imagination-Based Optimization\\ 
-  ​- ​EPOpt: Learning Robust Neural Network Policies Using Model Ensembles+C32: Recurrent Environment Simulators\\ 
 +C33: EPOpt: Learning Robust Neural Network Policies Using Model Ensembles\\
  
 +<​html><​div id='​wednesday_morning'></​div></​html>​
 ====Wednesday Morning (April 26th, 10:30am to 12:​30pm)==== ====Wednesday Morning (April 26th, 10:30am to 12:​30pm)====
-  - Deep Multi-task Representation Learning: A Tensor Factorisation Approach +C1: Deep Multi-task Representation Learning: A Tensor Factorisation Approach\\ 
-  ​- ​Training deep neural-networks using a noise adaptation layer +C2: Training deep neural-networks using a noise adaptation layer\\ 
-  ​- ​Delving into Transferable Adversarial Examples and Black-box Attacks +C3: Delving into Transferable Adversarial Examples and Black-box Attacks\\ 
-  ​- ​Towards the Limit of Network Quantization +C4: Towards the Limit of Network Quantization\\ 
-  ​- ​Towards Deep Interpretability (MUS-ROVER II): Learning Hierarchical Representations of Tonal Music +C5: Towards Deep Interpretability (MUS-ROVER II): Learning Hierarchical Representations of Tonal Music\\ 
-  ​- ​Learning to superoptimize programs +C6: Learning to superoptimize programs\\ 
-  ​- ​Regularizing CNNs with Locally Constrained Decorrelations +C7: Regularizing CNNs with Locally Constrained Decorrelations\\ 
-  ​- ​Generative Multi-Adversarial Networks +C8: Generative Multi-Adversarial Networks\\ 
-  ​- ​Visualizing Deep Neural Network Decisions: Prediction Difference Analysis +C9: Visualizing Deep Neural Network Decisions: Prediction Difference Analysis\\ 
-  ​- ​FractalNet: Ultra-Deep Neural Networks without Residuals +C10: FractalNet: Ultra-Deep Neural Networks without Residuals\\ 
-  ​- ​Faster CNNs with Direct Sparse Convolutions and Guided Pruning +C11: Faster CNNs with Direct Sparse Convolutions and Guided Pruning\\ 
-  ​- ​FILTER SHAPING FOR CONVOLUTIONAL NEURAL NETWORKS +C12: FILTER SHAPING FOR CONVOLUTIONAL NEURAL NETWORKS\\ 
-  ​- ​The Neural Noisy Channel +C13: The Neural Noisy Channel\\ 
-  ​- ​Automatic Rule Extraction from Long Short Term Memory Networks +C14: Automatic Rule Extraction from Long Short Term Memory Networks\\ 
-  ​- ​Adversarially Learned Inference +C15: Adversarially Learned Inference\\ 
-  ​- ​Deep Information Propagation +C16: Deep Information Propagation\\ 
-  ​- ​Revisiting Classifier Two-Sample Tests +C17: Revisiting Classifier Two-Sample Tests\\ 
-  ​- ​Loss-aware Binarization of Deep Networks +C18: Loss-aware Binarization of Deep Networks\\ 
-  ​- ​Energy-based Generative Adversarial Networks +C19: Energy-based Generative Adversarial Networks\\ 
-  ​- ​Central Moment Discrepancy (CMD) for Domain-Invariant Representation Learning +C20: Central Moment Discrepancy (CMD) for Domain-Invariant Representation Learning\\ 
-  ​- ​Temporal Ensembling for Semi-Supervised Learning +C21: Temporal Ensembling for Semi-Supervised Learning\\ 
-  ​- ​On Detecting Adversarial Perturbations +C22: On Detecting Adversarial Perturbations\\ 
-  - Identity Matters in Deep Learning +C23: Understanding deep learning requires rethinking generalization\\ 
-  ​- ​Adversarial Feature Learning +C24: Adversarial Feature Learning\\ 
-  ​- ​Learning through Dialogue Interactions +C25: Learning through Dialogue Interactions\\ 
-  ​- ​Learning to Compose Words into Sentences with Reinforcement Learning +C26: Learning to Compose Words into Sentences with Reinforcement Learning\\ 
-  ​- ​Batch Policy Gradient Methods for Improving Neural Conversation Models +C27: Batch Policy Gradient Methods for Improving Neural Conversation Models\\ 
-  ​- ​Tying Word Vectors and Word Classifiers:​ A Loss Framework for Language Modeling +C28: Tying Word Vectors and Word Classifiers:​ A Loss Framework for Language Modeling\\ 
-  ​- ​Geometry of Polysemy +C29: Geometry of Polysemy\\ 
-  ​- ​PGQ: Combining policy gradient and Q-learning +C30: PGQ: Combining policy gradient and Q-learning\\ 
-  ​- ​Reinforcement Learning through Asynchronous Advantage Actor-Critic on a GPU +C31: Reinforcement Learning through Asynchronous Advantage Actor-Critic on a GPU\\ 
-  ​- ​Learning to Navigate in Complex Environments +C32: Learning to Navigate in Complex Environments\\ 
-  ​- ​Learning and Policy Search in Stochastic Dynamical Systems with Bayesian Neural Networks+C33: Learning and Policy Search in Stochastic Dynamical Systems with Bayesian Neural Networks\\
  
 +<​html><​div id='​wednesday_afternoon'></​div></​html>​
 ====Wednesday Afternoon (April 26th, 4:30pm to 6:30pm)==== ====Wednesday Afternoon (April 26th, 4:30pm to 6:30pm)====
-  - Learning recurrent representations for hierarchical behavior modeling +C1: Learning recurrent representations for hierarchical behavior modeling\\ 
-  ​- ​Predicting Medications from Diagnostic Codes with Recurrent Neural Networks +C2: Predicting Medications from Diagnostic Codes with Recurrent Neural Networks\\ 
-  ​- ​Sparsely-Connected Neural Networks: Towards Efficient VLSI Implementation of Deep Neural Networks +C3: Sparsely-Connected Neural Networks: Towards Efficient VLSI Implementation of Deep Neural Networks\\ 
-  ​- ​HolStep: A Machine Learning Dataset for Higher-order Logic Theorem Proving +C4: HolStep: A Machine Learning Dataset for Higher-order Logic Theorem Proving\\ 
-  ​- ​Learning Invariant Representations Of Planar Curves +C5: Learning Invariant Representations Of Planar Curves\\ 
-  ​- ​Entropy-SGD:​ Biasing Gradient Descent Into Wide Valleys +C6: Entropy-SGD:​ Biasing Gradient Descent Into Wide Valleys\\ 
-  ​- ​Amortised MAP Inference for Image Super-resolution +C7: Amortised MAP Inference for Image Super-resolution\\ 
-  ​- ​Inductive Bias of Deep Convolutional Networks through Pooling Geometry +C8: Inductive Bias of Deep Convolutional Networks through Pooling Geometry\\ 
-  ​- ​Neural Photo Editing with Introspective Adversarial Networks +C9: Neural Photo Editing with Introspective Adversarial Networks\\ 
-  ​- ​A Learned Representation For Artistic Style +C10: A Learned Representation For Artistic Style\\ 
-  - Adversarial Machine ​Learning ​at Scale +C11: Learning ​to Remember Rare Events\\ 
-  Stick-Breaking Variational Autoencoders +C12: Optimization as a Model for Few-Shot Learning\\ 
-  ​- ​Support Regularized Sparse Coding and Its Fast Encoder +C13: Support Regularized Sparse Coding and Its Fast Encoder\\ 
-  ​- ​Discrete Variational Autoencoders +C14: Discrete Variational Autoencoders\\ 
-  Do Deep Convolutional Nets Really Need to be Deep and Convolutional?​ +C15: Training Compressed Fully-Connected Networks with a Density-Diversity Penalty\\ 
-  ​- ​Efficient Representation of Low-Dimensional Manifolds using Deep Networks +C16: Efficient Representation of Low-Dimensional Manifolds using Deep Networks\\ 
-  ​- ​Semi-Supervised Classification with Graph Convolutional Networks +C17: Semi-Supervised Classification with Graph Convolutional Networks\\ 
-  ​- ​Understanding Neural Sparse Coding with Matrix Factorization +C18: Understanding Neural Sparse Coding with Matrix Factorization\\ 
-  ​- ​Tighter bounds lead to improved classifiers +C19: Tighter bounds lead to improved classifiers\\ 
-  ​- ​Why Deep Neural Networks for Function Approximation?​ +C20: Why Deep Neural Networks for Function Approximation?​\\ 
-  ​- ​Hierarchical Multiscale Recurrent Neural Networks +C21: Hierarchical Multiscale Recurrent Neural Networks\\ 
-  ​- ​Dropout with Expectation-linear Regularization +C22: Dropout with Expectation-linear Regularization\\ 
-  ​- ​HyperNetworks +C23: HyperNetworks\\ 
-  ​- ​Hadamard Product for Low-rank Bilinear Pooling +C24: Hadamard Product for Low-rank Bilinear Pooling\\ 
-  ​- ​Adversarial Training Methods for Semi-Supervised Text Classification +C25: Adversarial Training Methods for Semi-Supervised Text Classification\\ 
-  ​- ​Fine-grained Analysis of Sentence Embeddings Using Auxiliary Prediction Tasks +C26: Fine-grained Analysis of Sentence Embeddings Using Auxiliary Prediction Tasks\\ 
-  ​- ​Pointer Sentinel Mixture Models +C27: Pointer Sentinel Mixture Models\\ 
-  ​- ​Reasoning with Memory Augmented Neural Networks for Language Comprehension +C28: Reasoning with Memory Augmented Neural Networks for Language Comprehension\\ 
-  ​- ​Dialogue Learning With Human-in-the-Loop +C29: Dialogue Learning With Human-in-the-Loop\\ 
-  - Learning to RepeatFine Grained Action Repetition for Deep Reinforcement Learning +C30Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations\\ 
-  ​- ​Learning to Play in a Day: Faster Deep Reinforcement Learning by Optimality Tightening +C31: Learning to Play in a Day: Faster Deep Reinforcement Learning by Optimality Tightening\\ 
-  ​- ​Learning Visual Servoing with Deep Features and Trust Region Fitted Q-Iteration +C32: Learning Visual Servoing with Deep Features and Trust Region Fitted Q-Iteration\\ 
-  ​- ​An Actor-Critic Algorithm for Sequence Prediction+C33: An Actor-Critic Algorithm for Sequence Prediction\\