Alert button
Picture for Joseph Najnudel

Joseph Najnudel

Alert button

Universal characteristics of deep neural network loss surfaces from random matrix theory

Add code
Bookmark button
Alert button
May 17, 2022
Nicholas P Baskerville, Jonathan P Keating, Francesco Mezzadri, Joseph Najnudel, Diego Granziol

Figure 1 for Universal characteristics of deep neural network loss surfaces from random matrix theory
Figure 2 for Universal characteristics of deep neural network loss surfaces from random matrix theory
Figure 3 for Universal characteristics of deep neural network loss surfaces from random matrix theory
Figure 4 for Universal characteristics of deep neural network loss surfaces from random matrix theory
Viaarxiv icon

A spin-glass model for the loss surfaces of generative adversarial networks

Add code
Bookmark button
Alert button
Jan 07, 2021
Nicholas P Baskerville, Jonathan P Keating, Francesco Mezzadri, Joseph Najnudel

Figure 1 for A spin-glass model for the loss surfaces of generative adversarial networks
Figure 2 for A spin-glass model for the loss surfaces of generative adversarial networks
Figure 3 for A spin-glass model for the loss surfaces of generative adversarial networks
Figure 4 for A spin-glass model for the loss surfaces of generative adversarial networks
Viaarxiv icon

The Loss Surfaces of Neural Networks with General Activation Functions

Add code
Bookmark button
Alert button
Apr 16, 2020
Nicholas P. Baskerville, Jonathan P. Keating, Francesco Mezzadri, Joseph Najnudel

Figure 1 for The Loss Surfaces of Neural Networks with General Activation Functions
Figure 2 for The Loss Surfaces of Neural Networks with General Activation Functions
Figure 3 for The Loss Surfaces of Neural Networks with General Activation Functions
Figure 4 for The Loss Surfaces of Neural Networks with General Activation Functions
Viaarxiv icon