There were a range of interesting papers at NIPS 2016 at Barcelona. Extension to adversarial networks were particularly prominent. DeepMath is really exciting as it is for the first time deep networks are utilized for large scale theorem proving.

​

Here are a few papers from a long list of very interesting papers presented this year at NIPS:

Feel free to add more in the comments. Enjoy!

​

• Human Decision-Making under Limited Time
   

• Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modeling
   

• Incremental Boosting Convolutional Neural Network for Facial Action Unit Recognition
   

• Tree-Structured Reinforcement Learning for Sequential Object Localization
   

• Blind Attacks on Machine Learners
   

• Variational Autoencoder for Deep Learning of Images, Labels and Captions
   

• Learning the Number of Neurons in Deep Networks
   

• DeepMath - Deep Sequence Models for Premise Selection
   

• Adaptive optimal training of animal behavior
   

• Improved Deep Metric Learning with Multi-class N-pair Loss Objective
   

• Correlated-PCA: Principal Components’ Analysis when Data and Noise are Correlated
   

• Doubly Convolutional Neural Networks
   

• PerforatedCNNs: Acceleration through Elimination of Redundant Convolutions
   

• Deep Alternative Neural Network: Exploring Contexts as Early as Possible for Action Recognition
   

• Learning feed-forward one-shot learners
   

• Domain Separation Networks
   

• Learning What and Where to Draw
   

• Professor Forcing: A New Algorithm for Training Recurrent Networks
   

• Adaptive Smoothed Online Multi-Task Learning
   

• Learning to learn by gradient descent by gradient descent
   

• Greedy Feature Construction
   

• Cooperative Inverse Reinforcement Learning
   

• Large-Scale Price Optimization via Network Flow
   

• CliqueCNN: Deep Unsupervised Exemplar Learning
   

• A Non-generative Framework and Convex Relaxations for Unsupervised Learning
   

• A Probabilistic Model of Social Decision Making based on Reward Maximization
   

• Generative Shape Models: Joint Text Recognition and Segmentation with Very Little Training Data
   

• Conditional Image Generation with PixelCNN Decoders

​

​

​