No article text was recorded for this version.
sparsity-aware weight initialization
Version 1 of 112 papers21 conceptsfrozen 2026-10-04
Papers in this version
- Blog: Gradient Flow in Sparse Neural Networks and How Lottery Tickets Win
- Blog: Robust Pruning at Initialization
- Blog: Training Your Sparse Neural Network Better with Any Mask
- Blog: Training Your Sparse Neural Network Better with Any Mask
- Blog: Signing the Supermask: Keep, Hide, Invert
- Blog: UltraMemV2: Memory Networks Scaling to 120B Parameters with Superior Long-Context Learning
- Blog: What's Hidden in a Randomly Weighted Neural Network?
- Blog: Language Models are Super Mario: Absorbing Abilities from Homologous Models as a Free Lunch
- Blog: ImPart: Importance-Aware Delta-Sparsification for Improved Model Compression and Merging in LLMs
- Blog: On the Neural Tangent Kernel Analysis of Randomly Pruned Neural Networks
- Blog: Epitopological learning and Cannistraci-Hebb network shape intelligence brain-inspired theory for ultra-sparse advantage in deep learning
- Blog: Butterfly Transform: An Efficient FFT Based Neural Architecture Design
History & citation
Current · Permanent link
- This direction split from Variance Compensation for Sparse Signal Propagation (92% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_sparsity_aware_weight_initialization_ee85d435,
title = {sparsity-aware weight initialization},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/sparsity-aware-weight-initialization/rdr_ee85d435b69a42bb9943fccca126c71e},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}This exact version (version 1, frozen text):
@misc{lacuna_sparsity_aware_weight_initialization_ee85d435_v1,
title = {sparsity-aware weight initialization},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/sparsity-aware-weight-initialization/rdr_ee85d435b69a42bb9943fccca126c71e/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}