← Search

Katie Everett

1 accepted papers

2025

Dimension-adapted Momentum Outscales SGD

NeurIPS 2025spotlight

We investigate scaling laws for stochastic momentum algorithms on the power law random features model, parameterized by data complexity, target complexity, and model size. When trained with a stochastic momentum algorithm, our analysis reveals four distinct loss curve shapes determined by varying da…

Cited by 0SourceScholar