**This is an old revision of the document!**

# Conference Poster Sessions

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.

### Monday Morning (April 24th, 10:30am to 12:30pm)

- Making Neural Programming Architectures Generalize via Recursion
- Learning Graphical State Transitions
- Distributed Second-Order Optimization using Kronecker-Factored Approximations
- Normalizing the Normalizers: Comparing and Extending Network Normalization Schemes
- Neural Program Lattices
- Diet Networks: Thin Parameters for Fat Genomics
- Unsupervised Cross-Domain Image Generation
- Towards Principled Methods for Training Generative Adversarial Networks
- Recurrent Mixture Density Network for Spatiotemporal Visual Attention
- Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfer
- Pruning Filters for Efficient ConvNets
- Optimization as a Model for Few-Shot Learning
- Understanding deep learning requires rethinking generalization
- On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima
- Recurrent Hidden Semi-Markov Model
- Nonparametric Neural Networks
- Learning to Generate Samples from Noise through Infusion Training
- An Information-Theoretic Framework for Fast and Robust Unsupervised Learning via Neural Population Infomax
- Highway and Residual Networks learn Unrolled Iterative Estimation
- Soft Weight-Sharing for Neural Network Compression
- Snapshot Ensembles: Train 1, Get M for Free
- Towards a Neural Statistician
- Learning Curve Prediction with Bayesian Neural Networks
- Learning End-to-End Goal-Oriented Dialog
- Multi-Agent Cooperation and the Emergence of (Natural) Language
- Efficient Vector Representation for Documents through Corruption
- Improving Neural Language Models with a Continuous Cache
- Program Synthesis for Character Level Language Modeling
- Tracking the World State with Recurrent Entity Networks
- Reinforcement Learning with Unsupervised Auxiliary Tasks
- Neural Architecture Search with Reinforcement Learning
- Sample Efficient Actor-Critic with Experience Replay
- Learning to Act by Predicting the Future

### Monday Afternoon (April 24th, 4:30pm to 6:30pm)

- Neuro-Symbolic Program Synthesis
- Generative Models and Model Criticism via Optimized Maximum Mean Discrepancy
- Trained Ternary Quantization
- DSD: Dense-Sparse-Dense Training for Deep Neural Networks
- A Compositional Object-Based Approach to Learning Physical Dynamics
- Multilayer Recurrent Network Models of Primate Retinal Ganglion Cells
- Improving Generative Adversarial Networks with Denoising Feature Matching
- Transfer of View-manifold Learning to Similarity Perception of Novel Objects
- What does it take to generate natural textures?
- Emergence of foveal image sampling from learning to attend in visual scenes
- PixelCNN++: A PixelCNN Implementation with Discretized Logistic Mixture Likelihood and Other Modifications
- Learning to Optimize
- Training Compressed Fully-Connected Networks with a Density-Diversity Penalty
- Optimal Binary Autoencoding with Pairwise Correlations
- On the Quantitative Analysis of Decoder-Based Generative Models
- Learning to Remember Rare Events
- Transfer Learning for Sequence Tagging with Hierarchical Recurrent Networks
- Capacity and Learnability in Recurrent Neural Networks
- Deep Learning with Dynamic Computation Graphs
- Exploring Sparsity in Recurrent Neural Networks
- Structured Attention Networks
- Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations
- Variational Lossy Autoencoder
- Learning to Query, Reason, and Answer Questions On Ambiguous Texts
- Deep Biaffine Attention for Neural Dependency Parsing
- A Compare-Aggregate Model for Matching Text Sequences
- Data Noising as Smoothing in Neural Network Language Models
- Neural Variational Inference For Topic Models
- Words or Characters? Fine-grained Gating for Reading Comprehension
- Q-Prop: Sample-Efficient Policy Gradient with An Off-Policy Critic
- Stochastic Neural Networks for Hierarchical Reinforcement Learning
- Learning Invariant Feature Spaces to Transfer Skills with Reinforcement Learning
- Third Person Imitation Learning

### Tuesday Morning (April 25th, 10:30am to 12:30pm)

- DeepDSL: A Compilation-based Domain-Specific Language for Deep Learning
- SampleRNN: An Unconditional End-to-End Neural Audio Generation Model
- Deep Probabilistic Programming
- Lie-Access Neural Turing Machines
- Learning Features of Music From Scratch
- Mode Regularized Generative Adversarial Networks
- End-to-end Optimized Image Compression
- Variational Recurrent Adversarial Deep Domain Adaptation
- Steerable CNNs
- Deep Predictive Coding Networks for Video Prediction and Unsupervised Learning
- PixelVAE: A Latent Variable Model for Natural Images
- A recurrent neural network without chaos
- Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer
- Tree-structured decoding with doubly-recurrent neural networks
- Introspection:Accelerating Neural Network Training By Learning Weight Evolution
- Hyperband: Bandit-Based Configuration Evaluation for Hyperparameter Optimization
- Quasi-Recurrent Neural Networks
- 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
- Trusting SVM for Piecewise Linear CNNs
- Maximum Entropy Flow Networks
- The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables
- Unrolled Generative Adversarial Networks
- A Simple but Tough-to-Beat Baseline for Sentence Embeddings
- Query-Reduction Networks for Question Answering
- Machine Comprehension Using Match-LSTM and Answer Pointer
- Bidirectional Attention Flow for Machine Comprehension
- Dynamic Coattention Networks For Question Answering
- Multi-view Recurrent Neural Acoustic Word Embeddings
- Episodic Exploration for Deep Deterministic Policies for StarCraft Micromanagement
- Training Agent for First-Person Shooter Game with Actor-Critic Curriculum Learning
- Generalizing Skills with Semi-Supervised Reinforcement Learning
- Improving Policy Gradient by Exploring Under-appreciated Rewards

### Tuesday Afternoon (April 25th, 4:30pm to 6:30pm)

- Sigma Delta Quantized Networks
- Paleo: A Performance Model for Deep Neural Networks
- DeepCoder: Learning to Write Programs
- Topology and Geometry of Deep Rectified Network Optimization Landscapes
- Incremental Network Quantization: Towards Lossless CNNs with Low-precision Weights
- Learning to Perform Physics Experiments via Deep Reinforcement Learning
- Decomposing Motion and Content for Natural Video Sequence Prediction
- Calibrating Energy-based Generative Adversarial Networks
- Pruning Convolutional Neural Networks for Resource Efficient Inference
- Incorporating long-range consistency in CNN-based texture generation
- Lossy Image Compression with Compressive Autoencoders
- LR-GAN: Layered Recursive Generative Adversarial Networks for Image Generation
- Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data
- Deep Variational Bayes Filters: Unsupervised Learning of State Space Models from Raw Data
- Mollifying Networks
- beta-VAE: Learning Basic Visual Concepts with a Constrained Variational Framework
- Categorical Reparameterization with Gumbel-Softmax
- Online Bayesian Transfer Learning for Sequential Data Modeling
- Latent Sequence Decompositions
- Density estimation using Real NVP
- Recurrent Batch Normalization
- SGDR: Stochastic Gradient Descent with Restarts
- Variable Computation in Recurrent Neural Networks
- Deep Variational Information Bottleneck
- A SELF-ATTENTIVE SENTENCE EMBEDDING
- TopicRNN: A Recurrent Neural Network with Long-Range Semantic Dependency
- Frustratingly Short Attention Spans in Neural Language Modeling
- Offline Bilingual Word Vectors Without a Dictionary
- LEARNING A NATURAL LANGUAGE INTERFACE WITH NEURAL PROGRAMMER
- Designing Neural Network Architectures using Reinforcement Learning
- Metacontrol for Adaptive Imagination-Based Optimization
- Recurrent Environment Simulators
- EPOpt: Learning Robust Neural Network Policies Using Model Ensembles

### Wednesday Morning (April 26th, 10:30am to 12:30pm)

- Deep Multi-task Representation Learning: A Tensor Factorisation Approach
- Training deep neural-networks using a noise adaptation layer
- Delving into Transferable Adversarial Examples and Black-box Attacks
- Towards the Limit of Network Quantization
- Towards Deep Interpretability (MUS-ROVER II): Learning Hierarchical Representations of Tonal Music
- Learning to superoptimize programs
- Regularizing CNNs with Locally Constrained Decorrelations
- Generative Multi-Adversarial Networks
- Visualizing Deep Neural Network Decisions: Prediction Difference Analysis
- FractalNet: Ultra-Deep Neural Networks without Residuals
- Faster CNNs with Direct Sparse Convolutions and Guided Pruning
- FILTER SHAPING FOR CONVOLUTIONAL NEURAL NETWORKS
- The Neural Noisy Channel
- Automatic Rule Extraction from Long Short Term Memory Networks
- Adversarially Learned Inference
- Deep Information Propagation
- Revisiting Classifier Two-Sample Tests
- Loss-aware Binarization of Deep Networks
- Energy-based Generative Adversarial Networks
- Central Moment Discrepancy (CMD) for Domain-Invariant Representation Learning
- Temporal Ensembling for Semi-Supervised Learning
- On Detecting Adversarial Perturbations
- Identity Matters in Deep Learning
- Adversarial Feature Learning
- Learning through Dialogue Interactions
- Learning to Compose Words into Sentences with Reinforcement Learning
- Batch Policy Gradient Methods for Improving Neural Conversation Models
- Tying Word Vectors and Word Classifiers: A Loss Framework for Language Modeling
- Geometry of Polysemy
- PGQ: Combining policy gradient and Q-learning
- Reinforcement Learning through Asynchronous Advantage Actor-Critic on a GPU
- Learning to Navigate in Complex Environments
- Learning and Policy Search in Stochastic Dynamical Systems with Bayesian Neural Networks

### Wednesday Afternoon (April 26th, 4:30pm to 6:30pm)

- Learning recurrent representations for hierarchical behavior modeling
- Predicting Medications from Diagnostic Codes with Recurrent Neural Networks
- Sparsely-Connected Neural Networks: Towards Efficient VLSI Implementation of Deep Neural Networks
- HolStep: A Machine Learning Dataset for Higher-order Logic Theorem Proving
- Learning Invariant Representations Of Planar Curves
- Entropy-SGD: Biasing Gradient Descent Into Wide Valleys
- Amortised MAP Inference for Image Super-resolution
- Inductive Bias of Deep Convolutional Networks through Pooling Geometry
- Neural Photo Editing with Introspective Adversarial Networks
- A Learned Representation For Artistic Style
- Adversarial Machine Learning at Scale
- Stick-Breaking Variational Autoencoders
- Support Regularized Sparse Coding and Its Fast Encoder
- Discrete Variational Autoencoders
- Do Deep Convolutional Nets Really Need to be Deep and Convolutional?
- Efficient Representation of Low-Dimensional Manifolds using Deep Networks
- Semi-Supervised Classification with Graph Convolutional Networks
- Understanding Neural Sparse Coding with Matrix Factorization
- Tighter bounds lead to improved classifiers
- Why Deep Neural Networks for Function Approximation?
- Hierarchical Multiscale Recurrent Neural Networks
- Dropout with Expectation-linear Regularization
- HyperNetworks
- Hadamard Product for Low-rank Bilinear Pooling
- Adversarial Training Methods for Semi-Supervised Text Classification
- Fine-grained Analysis of Sentence Embeddings Using Auxiliary Prediction Tasks
- Pointer Sentinel Mixture Models
- Reasoning with Memory Augmented Neural Networks for Language Comprehension
- Dialogue Learning With Human-in-the-Loop
- Learning to Repeat: Fine Grained Action Repetition for Deep Reinforcement Learning
- 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
- An Actor-Critic Algorithm for Sequence Prediction