No article text was recorded for this version.
identity-biased initialization
Version 1 of 120 papers28 conceptsfrozen 2026-10-04
Papers in this version
- Blog: Gradient descent with identity initialization efficiently learns positive definite linear transformations
- Blog: On the Global Convergence of Training Deep Linear ResNets
- Blog: Training Very Deep Networks
- Blog: Towards Understanding the Importance of Shortcut Connections in Residual Networks
- Blog: Bag of Tricks for Image Classification with Convolutional Neural Networks
- Blog: QuickSRNet: Plain Single-Image Super-Resolution Architecture for Faster Inference on Mobile Platforms
- Blog: The Shattered Gradients Problem: If resnets are the answer, then what is the question?
- Blog: Light Differentiable Logic Gate Networks
- Blog: Memorization and Optimization in Deep Neural Networks with Minimum Over-parameterization
- Blog: On the Optimization of Deep Networks: Implicit Acceleration by Overparameterization
- Blog: ACDC: A Structured Efficient Linear Layer Marcin Moczulski
- Blog: Residual Neural Networks for Digital Predistortion
- Blog: Gradient descent with identity initialization efficiently learns positive definite linear transformations by deep residual networks
- Blog: Residual Distillation: Towards Portable Deep Neural Networks without Shortcuts
- Blog: Deep linear networks for regression are implicitly regularized towards flat minima
- Blog: The Surprising Simplicity of the Early-Time Learning Dynamics of Neural Networks
- Blog: Spatially Adaptive Computation Time for Residual Networks
- Blog: Identity Matters in Deep Learning
- Blog: Neural Networks Learning and Memorization with (almost) no Over-Parameterization
- Blog: Gradients explode - Deep Networks are shallow - ResNet explained
History & citation
Current · Permanent link
Cite this direction
Latest version (a living page):
@misc{lacuna_identity_biased_initialization_46519cf8,
title = {identity-biased initialization},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/identity-biased-initialization/rdr_46519cf81d8e499b9af2b86a23dd14fa},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-22},
}This exact version (version 1, frozen text):
@misc{lacuna_identity_biased_initialization_46519cf8_v1,
title = {identity-biased initialization},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/identity-biased-initialization/rdr_46519cf81d8e499b9af2b86a23dd14fa/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-22},
}