NeurIPS 2017 accepted papers on Architectures and efficiency
57 of the 688 papers accepted at NeurIPS 2017 (8.3%) are on Architectures and efficiency, compared with 7.4% at NeurIPS 2016. They fall into 6 topics.
See NeurIPS 2017 in the interactive map
Submitting to NeurIPS? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
RNNs and normalization · 31 papers
- Self-Normalizing Neural Networks (511 citations)
- The Expressive Power of Neural Networks: A View from the Width (424 citations)
- Batch Renormalization: Towards Reducing Minibatch Dependence in Batch-Normalized Models (397 citations)
- Learned in Translation: Contextualized Word Vectors (237 citations)
- SVCCA: Singular Vector Canonical Correlation Analysis for Deep Learning Dynamics and Interpretability (218 citations)
- Learned D-AMP: Principled Neural Network based Compressive Image Recovery (189 citations)
- Gated Recurrent Convolution Neural Network for OCR (132 citations)
- Dilated Recurrent Neural Networks (123 citations)
- End-to-End Differentiable Proving (116 citations)
- Dynamic Routing Between Capsules (88 citations)
- Mean Field Residual Networks: On the Edge of Chaos (74 citations)
- Fast-Slow Recurrent Neural Networks (65 citations)
and 19 more in the interactive map.
CNN design · 11 papers
- Deep Learning for Precipitation Nowcasting: A Benchmark and A New Model (612 citations)
- Towards Accurate Binary Convolutional Neural Network (384 citations)
- The Reversible Residual Network: Backpropagation Without Storing Activations (351 citations)
- Working hard to know your neighbor's margins: Local descriptor learning loss (298 citations)
- Learning Spherical Convolution for Fast Features from 360° Imagery (131 citations)
- Learning Deep Structured Multi-Scale Features using Attention-Gated CRFs for Contour Prediction (103 citations)
- Deep Hyperspherical Learning (79 citations)
- Targeting EEG/LFP Synchrony with Neural Nets (50 citations)
- Neural system identification for large populations separating “what” and “where” (48 citations)
- Invariance and Stability of Deep Convolutional Representations (30 citations)
- Incorporating Side Information by Adaptive Convolution
Attention and transformers · 6 papers
- Attentional Pooling for Action Recognition (209 citations)
- Deconvolutional Paragraph Representation Learning (91 citations)
- Attend and Predict: Understanding Gene Regulation by Selective Attention on Chromatin (73 citations)
- Unbounded cache model for online language modeling with open vocabulary (11 citations)
- A Regularized Framework for Sparse and Structured Neural Attention
- Attention is All you Need
Pruning and efficient tuning · 4 papers
- Runtime Neural Pruning (355 citations)
- Compression-aware Training of Deep Networks (133 citations)
- Structured Bayesian Pruning via Log-Normal Multiplicative Noise (131 citations)
- Learning to Prune Deep Neural Networks via Layer-wise Optimal Brain Surgeon (63 citations)
Quantization · 4 papers
- Soft-to-Hard Vector Quantization for End-to-End Learning Compressible Representations (358 citations)
- Flexpoint: An Adaptive Numerical Format for Efficient Training of Deep Neural Networks (166 citations)
- Training Quantized Nets: A Deeper Understanding (141 citations)
- Simple strategies for recovering inner products from coarsely quantized random projections (8 citations)
Spiking networks · 1 paper
- Sparse convolutional coding for neuronal assembly detection (12 citations)
