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

# Workshop Poster Sessions

Below are the Workshop 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 Workshop Track.

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

- Extrapolation and learning equations
- Effectiveness of Transfer Learning in EHR data
- Intelligent synapses for multi-task and transfer learning
- Unsupervised and Efficient Neural Graph Model with Distributed Representations
- Accelerating SGD for Distributed Deep-Learning Using an Approximted Hessian Matrix
- Accelerating Eulerian Fluid Simulation With Convolutional Networks
- Forced to Learn: Discovering Disentangled Representations Without Exhaustive Labels
- Deep Nets Don't Learn via Memorization
- Learning Algorithms for Active Learning
- Reinterpreting Importance-Weighted Autoencoders
- Robustness to Adversarial Examples through an Ensemble of Specialists
- Neural Expectation Maximization
- On Hyperparameter Optimization in Learning Systems
- Recurrent Normalization Propagation
- Joint Training of Ratings and Reviews with Recurrent Recommender Networks
- Towards an Automatic Turing Test: Learning to Evaluate Dialogue Responses
- Joint Embeddings of Scene Graphs and Images
- Unseen Style Transfer Based on a Conditional Fast Style Transfer Network

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

- Audio Super-Resolution using Neural Networks
- Semantic embeddings for program behaviour patterns
- De novo drug design with deep generative models : an empirical study
- Memory Matching Networks for Genomic Sequence Classification
- Char2Wav: End-to-End Speech Synthesis
- Fast Chirplet Transform Injects Priors in Deep Learning of Animal Calls and Speech
- Weight-averaged consistency targets improve semi-supervised deep learning results
- Particle Value Functions
- Out-of-class novelty generation: an experimental foundation
- Performance guarantees for transferring representations
- Generative Adversarial Learning of Markov Chains
- Short and Deep: Sketching and Neural Networks
- Understanding intermediate layers using linear classifier probes
- Symmetry-Breaking Convergence Analysis of Certain Two-layered Neural Networks with ReLU nonlinearity
- Neural Combinatorial Optimization with Reinforcement Learning
- Tactics of Adversarial Attacks on Deep Reinforcement Learning Agents
- Adversarial Discriminative Domain Adaptation (workshop extended abstract)
- Efficient Sparse-Winograd Convolutional Neural Networks

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

- Programming With a Differentiable Forth Interpreter
- Unsupervised Feature Learning for Audio Analysis
- Neural Functional Programming
- A Smooth Optimisation Perspective on Training Feedforward Neural Networks
- Synthetic Gradient Methods with Virtual Forward-Backward Networks
- Explaining the Learning Dynamics of Direct Feedback Alignment
- Training a Subsampling Mechanism in Expectation
- Deep Kernel Machines via the Kernel Reparametrization Trick
- Encoding and Decoding Representations with Sum- and Max-Product Networks
- Embracing Data Abundance
- Variational Intrinsic Control
- Fast Adaptation in Generative Models with Generative Matching Networks
- Efficient variational Bayesian neural network ensembles for outlier detection
- Emergence of Language with Multi-agent Games: Learning to Communicate with Sequences of Symbols
- Adaptive Feature Abstraction for Translating Video to Language
- Delving into adversarial attacks on deep policies
- Tuning Recurrent Neural Networks with Reinforcement Learning
- DeepMask: Masking DNN Models for robustness against adversarial samples
- Restricted Boltzmann Machines provide an accurate metric for retinal responses to visual stimuli

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

- Lifelong Perceptual Programming By Example
- Neu0
- Dance Dance Convolution
- Bit-Pragmatic Deep Neural Network Computing
- On Improving the Numerical Stability of Winograd Convolutions
- Fast Generation for Convolutional Autoregressive Models
- THE PREIMAGE OF RECTIFIER NETWORK ACTIVITIES
- Training Triplet Networks with GAN
- On Robust Concepts and Small Neural Nets
- Pl@ntNet app in the era of deep learning
- Exponential Machines
- Online Multi-Task Learning Using Biased Sampling
- Online Structure Learning for Sum-Product Networks with Gaussian Leaves
- A Theoretical Framework for Robustness of (Deep) Classifiers against Adversarial Samples
- Compositional Kernel Machines
- Loss is its own Reward: Self-Supervision for Reinforcement Learning
- Changing Model Behavior at Test-time Using Reinforcement Learning
- Precise Recovery of Latent Vectors from Generative Adversarial Networks
- Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization

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

- NEUROGENESIS-INSPIRED DICTIONARY LEARNING: ONLINE MODEL ADAPTION IN A CHANGING WORLD
- The High-Dimensional Geometry of Binary Neural Networks
- Discovering objects and their relations from entangled scene representations
- A Differentiable Physics Engine for Deep Learning in Robotics
- Automated Generation of Multilingual Clusters for the Evaluation of Distributed Representations
- Development of JavaScript-based deep learning platform and application to distributed training
- Factorization tricks for LSTM networks
- Shake-Shake regularization of 3-branch residual networks
- Trace Norm Regularised Deep Multi-Task Learning
- Deep Learning with Sets and Point Clouds
- Dataset Augmentation in Feature Space
- Multiplicative LSTM for sequence modelling
- Learning to Discover Sparse Graphical Models
- Revisiting Batch Normalization For Practical Domain Adaptation
- Early Methods for Detecting Adversarial Images and a Colorful Saliency Map
- Natural Language Generation in Dialogue using Lexicalized and Delexicalized Data
- Coupling Distributed and Symbolic Execution for Natural Language Queries
- Adversarial Examples for Semantic Image Segmentation
- RenderGAN: Generating Realistic Labeled Data

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

- Song From PI: A Musically Plausible Network for Pop Music Generation
- Charged Point Normalization: An Efficient Solution to the Saddle Point Problem
- Towards “AlphaChem”: Chemical Synthesis Planning with Tree Search and Deep Neural Network Policies
- CommAI: Evaluating the first steps towards a useful general AI
- Joint Multimodal Learning with Deep Generative Models
- Transferring Knowledge to Smaller Network with Class-Distance Loss
- Regularizing Neural Networks by Penalizing Confident Output Distributions
- Adversarial Attacks on Neural Network Policies
- Generalizable Features From Unsupervised Learning
- Compact Embedding of Binary-coded Inputs and Outputs using Bloom Filters
- Semi-supervised deep learning by metric embedding
- REBAR: Low-variance, unbiased gradient estimates for discrete latent variable models
- Variational Reference Priors
- Gated Multimodal Units for Information Fusion
- Playing SNES in the Retro Learning Environment
- Unsupervised Perceptual Rewards for Imitation Learning
- Perception Updating Networks: On architectural constraints for interpretable video generative models
- Adversarial examples in the physical world