Timezone: »
It is well known that weight initialization in deep networks can have a dramatic impact on learning speed. For example, ensuring the mean squared singular value of a network's input-output Jacobian is O(1) is essential for avoiding exponentially vanishing or exploding gradients. Moreover, in deep linear networks, ensuring that all singular values of the Jacobian are concentrated near 1 can yield a dramatic additional speed-up in learning; this is a property known as dynamical isometry. However, it is unclear how to achieve dynamical isometry in nonlinear deep networks. We address this question by employing powerful tools from free probability theory to analytically compute the {\it entire} singular value distribution of a deep network's input-output Jacobian. We explore the dependence of the singular value distribution on the depth of the network, the weight initialization, and the choice of nonlinearity. Intriguingly, we find that ReLU networks are incapable of dynamical isometry. On the other hand, sigmoidal networks can achieve isometry, but only with orthogonal weight initialization. Moreover, we demonstrate empirically that deep nonlinear networks achieving dynamical isometry learn orders of magnitude faster than networks that do not. Indeed, we show that properly-initialized deep sigmoidal networks consistently outperform deep ReLU networks. Overall, our analysis reveals that controlling the entire distribution of Jacobian singular values is an important design consideration in deep learning.
Author Information
Jeffrey Pennington (Google Brain)
Samuel Schoenholz (Google Brain)
Surya Ganguli (Stanford)
More from the Same Authors
-
2020 : End-to-End Differentiability and Tensor Processing Unit Computing to Accelerate Materials’ Inverse Design »
HAN LIU · Yuhan Liu · Zhangji Zhao · Samuel Schoenholz · Ekin Dogus Cubuk · Mathieu Bauchy -
2021 : Fast Finite Width Neural Tangent Kernel »
Roman Novak · Jascha Sohl-Dickstein · Samuel Schoenholz -
2022 : Unmasking the Lottery Ticket Hypothesis: Efficient Adaptive Pruning for Finding Winning Tickets »
Mansheej Paul · Feng Chen · Brett Larsen · Jonathan Frankle · Surya Ganguli · Gintare Karolina Dziugaite -
2022 : A Second-order Regression Model Shows Edge of Stability Behavior »
Fabian Pedregosa · Atish Agarwala · Jeffrey Pennington -
2022 Poster: Lottery Tickets on a Data Diet: Finding Initializations with Sparse Trainable Networks »
Mansheej Paul · Brett Larsen · Surya Ganguli · Jonathan Frankle · Gintare Karolina Dziugaite -
2022 Poster: Implicit Regularization or Implicit Conditioning? Exact Risk Trajectories of SGD in High Dimensions »
Courtney Paquette · Elliot Paquette · Ben Adlam · Jeffrey Pennington -
2022 Poster: Precise Learning Curves and Higher-Order Scalings for Dot-product Kernel Regression »
Lechao Xiao · Hong Hu · Theodor Misiakiewicz · Yue Lu · Jeffrey Pennington -
2022 Poster: Beyond neural scaling laws: beating power law scaling via data pruning »
Ben Sorscher · Robert Geirhos · Shashank Shekhar · Surya Ganguli · Ari Morcos -
2021 : Session 3 | Invited talk: Surya Ganguli, "From the geometry of high dimensional energy landscapes to optimal annealing in a dissipative many body quantum optimizer" »
Surya Ganguli · Atilim Gunes Baydin -
2021 Poster: Deep Learning on a Data Diet: Finding Important Examples Early in Training »
Mansheej Paul · Surya Ganguli · Gintare Karolina Dziugaite -
2021 Poster: Overparameterization Improves Robustness to Covariate Shift in High Dimensions »
Nilesh Tripuraneni · Ben Adlam · Jeffrey Pennington -
2020 Poster: Deep learning versus kernel learning: an empirical study of loss landscape geometry and the time evolution of the Neural Tangent Kernel »
Stanislav Fort · Gintare Karolina Dziugaite · Mansheej Paul · Sepideh Kharaghani · Daniel Roy · Surya Ganguli -
2020 Poster: Finite Versus Infinite Neural Networks: an Empirical Study »
Jaehoon Lee · Samuel Schoenholz · Jeffrey Pennington · Ben Adlam · Lechao Xiao · Roman Novak · Jascha Sohl-Dickstein -
2020 Spotlight: Finite Versus Infinite Neural Networks: an Empirical Study »
Jaehoon Lee · Samuel Schoenholz · Jeffrey Pennington · Ben Adlam · Lechao Xiao · Roman Novak · Jascha Sohl-Dickstein -
2020 Poster: Predictive coding in balanced neural networks with noise, chaos and delays »
Jonathan Kadmon · Jonathan Timcheck · Surya Ganguli -
2020 Poster: Identifying Learning Rules From Neural Network Observables »
Aran Nayebi · Sanjana Srivastava · Surya Ganguli · Daniel Yamins -
2020 Poster: The Surprising Simplicity of the Early-Time Learning Dynamics of Neural Networks »
Wei Hu · Lechao Xiao · Ben Adlam · Jeffrey Pennington -
2020 Spotlight: The Surprising Simplicity of the Early-Time Learning Dynamics of Neural Networks »
Wei Hu · Lechao Xiao · Ben Adlam · Jeffrey Pennington -
2020 Spotlight: Identifying Learning Rules From Neural Network Observables »
Aran Nayebi · Sanjana Srivastava · Surya Ganguli · Daniel Yamins -
2020 Poster: JAX MD: A Framework for Differentiable Physics »
Samuel Schoenholz · Ekin Dogus Cubuk -
2020 Spotlight: JAX MD: A Framework for Differentiable Physics »
Samuel Schoenholz · Ekin Dogus Cubuk -
2020 Poster: Understanding Double Descent Requires A Fine-Grained Bias-Variance Decomposition »
Ben Adlam · Jeffrey Pennington -
2020 Poster: Pruning neural networks without any data by iteratively conserving synaptic flow »
Hidenori Tanaka · Daniel Kunin · Daniel Yamins · Surya Ganguli -
2019 : Panel Session: A new hope for neuroscience »
Yoshua Bengio · Blake Richards · Timothy Lillicrap · Ila Fiete · David Sussillo · Doina Precup · Konrad Kording · Surya Ganguli -
2019 : Poster Session »
Pravish Sainath · Mohamed Akrout · Charles Delahunt · Nathan Kutz · Guangyu Robert Yang · Joseph Marino · L F Abbott · Nicolas Vecoven · Damien Ernst · andrew warrington · Michael Kagan · Kyunghyun Cho · Kameron Harris · Leopold Grinberg · John J. Hopfield · Dmitry Krotov · Taliah Muhammad · Erick Cobos · Edgar Walker · Jacob Reimer · Andreas Tolias · Alexander Ecker · Janaki Sheth · Yu Zhang · Maciej Wołczyk · Jacek Tabor · Szymon Maszke · Roman Pogodin · Dane Corneil · Wulfram Gerstner · Baihan Lin · Guillermo Cecchi · Jenna M Reinen · Irina Rish · Guillaume Bellec · Darjan Salaj · Anand Subramoney · Wolfgang Maass · Yueqi Wang · Ari Pakman · Jin Hyung Lee · Liam Paninski · Bryan Tripp · Colin Graber · Alex Schwing · Luke Prince · Gabriel Ocker · Michael Buice · Benjamin Lansdell · Konrad Kording · Jack Lindsey · Terrence Sejnowski · Matthew Farrell · Eric Shea-Brown · Nicolas Farrugia · Victor Nepveu · Jiwoong Im · Kristin Branson · Brian Hu · Ramakrishnan Iyer · Stefan Mihalas · Sneha Aenugu · Hananel Hazan · Sihui Dai · Tan Nguyen · Doris Tsao · Richard Baraniuk · Anima Anandkumar · Hidenori Tanaka · Aran Nayebi · Stephen Baccus · Surya Ganguli · Dean Pospisil · Eilif Muller · Jeffrey S Cheng · Gaël Varoquaux · Kamalaker Dadi · Dimitrios C Gklezakos · Rajesh PN Rao · Anand Louis · Christos Papadimitriou · Santosh Vempala · Naganand Yadati · Daniel Zdeblick · Daniela M Witten · Nicholas Roberts · Vinay Prabhu · Pierre Bellec · Poornima Ramesh · Jakob H Macke · Santiago Cadena · Guillaume Bellec · Franz Scherr · Owen Marschall · Robert Kim · Hannes Rapp · Marcio Fonseca · Oliver Armitage · Jiwoong Im · Thomas Hardcastle · Abhishek Sharma · Wyeth Bair · Adrian Valente · Shane Shang · Merav Stern · Rutuja Patil · Peter Wang · Sruthi Gorantla · Peter Stratton · Tristan Edwards · Jialin Lu · Martin Ester · Yurii Vlasov · Siavash Golkar -
2019 : Panel - The Role of Communication at Large: Aparna Lakshmiratan, Jason Yosinski, Been Kim, Surya Ganguli, Finale Doshi-Velez »
Aparna Lakshmiratan · Finale Doshi-Velez · Surya Ganguli · Zachary Lipton · Michela Paganini · Anima Anandkumar · Jason Yosinski -
2019 : Afternoon Coffee Break & Poster Session »
Heidi Komkov · Stanislav Fort · Zhaoyou Wang · Rose Yu · Ji Hwan Park · Samuel Schoenholz · Taoli Cheng · Ryan-Rhys Griffiths · Chase Shimmin · Surya Karthik Mukkavili · Philippe Schwaller · Christian Knoll · Yangzesheng Sun · Keiichi Kisamori · Gavin Graham · Gavin Portwood · Hsin-Yuan Huang · Paul Novello · Moritz Munchmeyer · Anna Jungbluth · Daniel Levine · Ibrahim Ayed · Steven Atkinson · Jan Hermann · Peter Grönquist · · Priyabrata Saha · Yannik Glaser · Lingge Li · Yutaro Iiyama · Rushil Anirudh · Maciej Koch-Janusz · Vikram Sundar · Francois Lanusse · Auralee Edelen · Jonas Köhler · Jacky H. T. Yip · jiadong guo · Xiangyang Ju · Adi Hanuka · Adrian Albert · Valentina Salvatelli · Mauro Verzetti · Javier Duarte · Eric Moreno · Emmanuel de Bézenac · Athanasios Vlontzos · Alok Singh · Thomas Klijnsma · Brad Neuberg · Paul Wright · Mustafa Mustafa · David Schmidt · Steven Farrell · Hao Sun -
2019 : Invited Talk: Theories for the emergence of internal representations in neural networks: from perception to navigation »
Surya Ganguli -
2019 : Lunch Break and Posters »
Xingyou Song · Elad Hoffer · Wei-Cheng Chang · Jeremy Cohen · Jyoti Islam · Yaniv Blumenfeld · Andreas Madsen · Jonathan Frankle · Sebastian Goldt · Satrajit Chatterjee · Abhishek Panigrahi · Alex Renda · Brian Bartoldson · Israel Birhane · Aristide Baratin · Niladri Chatterji · Roman Novak · Jessica Forde · YiDing Jiang · Yilun Du · Linara Adilova · Michael Kamp · Berry Weinstein · Itay Hubara · Tal Ben-Nun · Torsten Hoefler · Daniel Soudry · Hsiang-Fu Yu · Kai Zhong · Yiming Yang · Inderjit Dhillon · Jaime Carbonell · Yanqing Zhang · Dar Gilboa · Johannes Brandstetter · Alexander R Johansen · Gintare Karolina Dziugaite · Raghav Somani · Ari Morcos · Freddie Kalaitzis · Hanie Sedghi · Lechao Xiao · John Zech · Muqiao Yang · Simran Kaur · Qianli Ma · Yao-Hung Hubert Tsai · Ruslan Salakhutdinov · Sho Yaida · Zachary Lipton · Daniel Roy · Michael Carbin · Florent Krzakala · Lenka Zdeborová · Guy Gur-Ari · Ethan Dyer · Dilip Krishnan · Hossein Mobahi · Samy Bengio · Behnam Neyshabur · Praneeth Netrapalli · Kris Sankaran · Julien Cornebise · Yoshua Bengio · Vincent Michalski · Samira Ebrahimi Kahou · Md Rifat Arefin · Jiri Hron · Jaehoon Lee · Jascha Sohl-Dickstein · Samuel Schoenholz · David Schwab · Dongyu Li · Sang Keun Choe · Henning Petzka · Ashish Verma · Zhichao Lin · Cristian Sminchisescu -
2019 : JAX, M.D.: End-to-End Differentiable, Hardware Accelerated, Molecular Dynamics in Pure Python »
Samuel Schoenholz -
2019 : Surya Ganguli, Yasaman Bahri, Florent Krzakala moderated by Lenka Zdeborova »
Florent Krzakala · Yasaman Bahri · Surya Ganguli · Lenka Zdeborová · Adji Bousso Dieng · Joan Bruna -
2019 : Surya Ganguli - An analytic theory of generalization dynamics and transfer learning in deep linear networks »
Surya Ganguli -
2019 Poster: Wide Neural Networks of Any Depth Evolve as Linear Models Under Gradient Descent »
Jaehoon Lee · Lechao Xiao · Samuel Schoenholz · Yasaman Bahri · Roman Novak · Jascha Sohl-Dickstein · Jeffrey Pennington -
2019 Poster: A unified theory for the origin of grid cells through the lens of pattern formation »
Ben Sorscher · Gabriel Mel · Surya Ganguli · Samuel Ocko -
2019 Poster: Universality and individuality in neural dynamics across large populations of recurrent networks »
Niru Maheswaranathan · Alex Williams · Matthew Golub · Surya Ganguli · David Sussillo -
2019 Spotlight: A unified theory for the origin of grid cells through the lens of pattern formation »
Ben Sorscher · Gabriel Mel · Surya Ganguli · Samuel Ocko -
2019 Spotlight: Universality and individuality in neural dynamics across large populations of recurrent networks »
Niru Maheswaranathan · Alex Williams · Matthew Golub · Surya Ganguli · David Sussillo -
2019 Poster: MetaInit: Initializing learning by learning to initialize »
Yann Dauphin · Samuel Schoenholz -
2019 Poster: From deep learning to mechanistic understanding in neuroscience: the structure of retinal prediction »
Hidenori Tanaka · Aran Nayebi · Niru Maheswaranathan · Lane McIntosh · Stephen Baccus · Surya Ganguli -
2019 Poster: Reverse engineering recurrent networks for sentiment classification reveals line attractor dynamics »
Niru Maheswaranathan · Alex Williams · Matthew Golub · Surya Ganguli · David Sussillo -
2018 Poster: The emergence of multiple retinal cell types through efficient coding of natural movies »
Samuel Ocko · Jack Lindsey · Surya Ganguli · Stephane Deny -
2018 Poster: The Spectrum of the Fisher Information Matrix of a Single-Hidden-Layer Neural Network »
Jeffrey Pennington · Pratik Worah -
2018 Poster: Statistical mechanics of low-rank tensor decomposition »
Jonathan Kadmon · Surya Ganguli -
2018 Poster: Task-Driven Convolutional Recurrent Models of the Visual System »
Aran Nayebi · Daniel Bear · Jonas Kubilius · Kohitij Kar · Surya Ganguli · David Sussillo · James J DiCarlo · Daniel Yamins -
2017 Spotlight: Nonlinear random matrix theory for deep learning »
Jeffrey Pennington · Pratik Worah -
2017 Poster: Nonlinear random matrix theory for deep learning »
Jeffrey Pennington · Pratik Worah -
2017 Poster: Variational Walkback: Learning a Transition Operator as a Stochastic Recurrent Net »
Anirudh Goyal · Nan Rosemary Ke · Surya Ganguli · Yoshua Bengio -
2017 Poster: Mean Field Residual Networks: On the Edge of Chaos »
Ge Yang · Samuel Schoenholz -
2016 : Surya Ganguli : Deep Neural Models of the Retinal Response to Natural Stimuli »
Surya Ganguli -
2016 : Non-convexity in the error landscape and the expressive capacity of deep neural networks »
Surya Ganguli -
2016 Poster: Exponential expressivity in deep neural networks through transient chaos »
Ben Poole · Subhaneil Lahiri · Maithra Raghu · Jascha Sohl-Dickstein · Surya Ganguli -
2016 Poster: An equivalence between high dimensional Bayes optimal inference and M-estimation »
Madhu Advani · Surya Ganguli -
2016 Poster: Deep Learning Models of the Retinal Response to Natural Scenes »
Lane McIntosh · Niru Maheswaranathan · Aran Nayebi · Surya Ganguli · Stephen Baccus -
2015 Poster: Spherical Random Features for Polynomial Kernels »
Jeffrey Pennington · Felix Yu · Sanjiv Kumar -
2015 Spotlight: Spherical Random Features for Polynomial Kernels »
Jeffrey Pennington · Felix Yu · Sanjiv Kumar -
2015 Poster: Deep Knowledge Tracing »
Chris Piech · Jonathan Bassen · Jonathan Huang · Surya Ganguli · Mehran Sahami · Leonidas Guibas · Jascha Sohl-Dickstein -
2014 Workshop: Deep Learning and Representation Learning »
Andrew Y Ng · Yoshua Bengio · Adam Coates · Roland Memisevic · Sharanyan Chetlur · Geoffrey E Hinton · Shamim Nemati · Bryan Catanzaro · Surya Ganguli · Herbert Jaeger · Phil Blunsom · Leon Bottou · Volodymyr Mnih · Chen-Yu Lee · Rich M Schwartz -
2014 Poster: Identifying and attacking the saddle point problem in high-dimensional non-convex optimization »
Yann N Dauphin · Razvan Pascanu · Caglar Gulcehre · Kyunghyun Cho · Surya Ganguli · Yoshua Bengio -
2013 Poster: A memory frontier for complex synapses »
Subhaneil Lahiri · Surya Ganguli -
2013 Oral: A memory frontier for complex synapses »
Subhaneil Lahiri · Surya Ganguli -
2010 Poster: Short-term memory in neuronal networks through dynamical compressed sensing »
Surya Ganguli · Haim Sompolinsky