model · first seen 5 Oct 2026
diagonal linear networks
Simplified linear networks with diagonal weight parameterization used to study implicit bias of gradient descent.
Mentions per week
PapersCodeDiscussion
Earliest sources
- Mind the Drift: Diagonal Linear Networks Under Large Learning RatesPapers · 5 Oct 2026 · Aniket Sanyal, Tom Jacobs, Rebekka Burkholz
- Continual Classification Beyond the Kernel Regime: Harnessing the Sparsity Bias of L2 Regularization in Diagonal Linear NetworksPapers · 4 Nov 2026 · Gilad Karpel, Itay Evron, Ron Meir, Daniel Soudry
Latest sources
- Continual Classification Beyond the Kernel Regime: Harnessing the Sparsity Bias of L2 Regularization in Diagonal Linear NetworksPapers · 4 Nov 2026 · Gilad Karpel, Itay Evron, Ron Meir, Daniel Soudry
- Mind the Drift: Diagonal Linear Networks Under Large Learning RatesPapers · 5 Oct 2026 · Aniket Sanyal, Tom Jacobs, Rebekka Burkholz