(ICML 2026) Balancing Learning Rates Across Layers: Exact Two-Step Dynamics and Optimal Scaling in Linear Neural Networks

Name		Name	Last commit message	Last commit date
Latest commit History 6 Commits
2-NN-Gaussian-experiment.py		2-NN-Gaussian-experiment.py
2-NN-Orthogonal-experiment.py		2-NN-Orthogonal-experiment.py
2-NN-Orthogonal-theory.py		2-NN-Orthogonal-theory.py
3-NN-Gaussian-experiment.py		3-NN-Gaussian-experiment.py
3-NN-Orthogonal-experiment.py		3-NN-Orthogonal-experiment.py
3-NN-Orthogonal-theory.py		3-NN-Orthogonal-theory.py
4-NN-Orthogonal-experiment.py		4-NN-Orthogonal-experiment.py
8-NN-Orthogonal-experiment.py		8-NN-Orthogonal-experiment.py
README.md		README.md

Provide feedback