Hyperball Optimizer Boosts Pretraining Speed by 20-30% When Combined with Muon

burny_tech · x · 2026-08-12

A new paper introduces Hyperball, an optimizer wrapper designed to address the issue of matrix-based optimizers like Muon losing their performance gains as model and data scales increase.

Core Mechanism:

Results:

Related event: Hyperball Optimizer Breaks Muon Scaling Bottleneck for 30% Speedup(2 posts)→

Original post →

More from Research

Research channel →