439 citations · 1.2k across the 10 of their papers we have counts for
8 papers · 1 filter
Top-KAST: Top-K Always Sparse Training
Siddhant M. Jayakumar, Razvan Pascanu, Jack W. Rae +2
Sparse neural networks are becoming increasingly important as the field seeks to improve the performance of existing models by scaling them up, while simultaneously trying to reduc…
On the Generalization Benefit of Noise in Stochastic Gradient Descent
Samuel L. Smith, Erich Elsen, Soham De
It has long been argued that minibatch stochastic gradient descent can generalize better than large batch gradient descent in deep neural networks. However recent papers have quest…
AlgebraNets
Jordan Hoffmann, Simon Schmitt, Simon Osindero +2
Neural networks have historically been built layerwise from the set of functions in , i.e. with activations and weights/parameters represented…
A Practical Sparse Approximation for Real Time Recurrent Learning
Jacob Menick, Erich Elsen, Utku Evci +3
Current methods for training recurrent neural networks are based on backpropagation through time, which requires storing a complete history of network states, and prohibits updatin…
Sparse GPU Kernels for Deep Learning
Trevor Gale, Matei Zaharia, Cliff Young +1
Scientific workloads have traditionally exploited high levels of sparsity to accelerate computation and reduce memory requirements. While deep neural networks can be made sparse, a…
The Difficulty of Training Sparse Neural Networks
Utku Evci, Fabian Pedregosa, Aidan Gomez +1
We investigate the difficulties of training sparse neural networks and make new observations about optimization dynamics and the energy landscape within the sparse regime. Recent w…