pith. sign in

The surprising agreement between convex optimization theory and learning-rate scheduling for large model training

2 Pith papers cite this work. Polarity classification is still indexing.

2 Pith papers citing it

fields

cs.LG 2

years

2026 2

verdicts

UNVERDICTED 2

representative citing papers

Optimistic Dual Averaging Unifies Modern Optimizers

cs.LG · 2026-05-11 · unverdicted · novelty 6.0

SODA unifies several modern optimizers under optimistic dual averaging and supplies a 1/k decay wrapper that improves performance without weight decay tuning.

Anytime Training with Schedule-Free Spectral Optimization

cs.LG · 2026-05-21 · unverdicted · novelty 5.0

SF-NorMuon is a new schedule-free spectral optimizer that closes the gap with tuned AdamW on 125M-772M parameter models across 1-8x Chinchilla horizons while providing stationarity guarantees.

citing papers explorer

Showing 2 of 2 citing papers.

  • Optimistic Dual Averaging Unifies Modern Optimizers cs.LG · 2026-05-11 · unverdicted · none · ref 15

    SODA unifies several modern optimizers under optimistic dual averaging and supplies a 1/k decay wrapper that improves performance without weight decay tuning.

  • Anytime Training with Schedule-Free Spectral Optimization cs.LG · 2026-05-21 · unverdicted · none · ref 56

    SF-NorMuon is a new schedule-free spectral optimizer that closes the gap with tuned AdamW on 125M-772M parameter models across 1-8x Chinchilla horizons while providing stationarity guarantees.