How to Start Training: The Effect of Initialization and Architecture

5 March 2018

Papers citing "How to Start Training: The Effect of Initialization and Architecture"

10 / 60 papers shown

Title
Optimization for deep learning: theory and algorithms Ruoyu Sun ODL 27 168 0 19 Dec 2019
Tensor Programs I: Wide Feedforward or Recurrent Neural Networks of Any Architecture are Gaussian Processes Greg Yang 33 193 0 28 Oct 2019
Non-Gaussian processes and neural networks at finite widths Sho Yaida 39 87 0 30 Sep 2019
Finite Depth and Width Corrections to the Neural Tangent Kernel Boris Hanin Mihai Nica MDE 30 149 0 13 Sep 2019
One ticket to win them all: generalizing lottery ticket initializations across datasets and optimizers Ari S. Morcos Haonan Yu Michela Paganini Yuandong Tian 16 228 0 06 Jun 2019
Infinitely deep neural networks as diffusion processes Stefano Peluchetti Stefano Favaro ODL 14 31 0 27 May 2019
Data driven approximation of parametrized PDEs by Reduced Basis and Neural Networks N. D. Santo S. Deparis Luca Pegolotti 19 66 0 02 Apr 2019
On the security relevance of weights in deep learning Kathrin Grosse T. A. Trost Marius Mosbach Michael Backes Dietrich Klakow AAML 32 6 0 08 Feb 2019
Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processes Roman Novak Lechao Xiao Jaehoon Lee Yasaman Bahri Greg Yang Jiri Hron Daniel A. Abolafia Jeffrey Pennington Jascha Narain Sohl-Dickstein UQCV BDL 25 307 0 11 Oct 2018
Dynamical Isometry and a Mean Field Theory of CNNs: How to Train 10,000-Layer Vanilla Convolutional Neural Networks Lechao Xiao Yasaman Bahri Jascha Narain Sohl-Dickstein S. Schoenholz Jeffrey Pennington 244 349 0 14 Jun 2018