After a fix, MuonM edges out Muon while MuonS performs about the same
YouJiacheng · x · 2026-11-19
YouJiacheng points to corrected experiment results on Muon optimizer variants: after the fix, MuonM performed slightly better than Muon, while MuonS was almost identical to the original. A useful empirical update for optimizer research and pretraining reproduction.
More from Research
- GPT 5.6 Sol, released just 4 months ago, can no longer reliably monitor frontier models — connoraxiotes · 2026-11-19
- Ambient SFT: treating the pre-finetune model as a noisy teacher, up to 21% gains — giannis_daras · 2026-11-19
- SREGym 2.0 recreates the Microsoft Outlook outage: a latent gateway bug plus cascading failures frontier AI can't yet solve — ajratner · 2026-11-19
- 'Naturals + infinity' form a nonstandard model of Robinson arithmetic — jessi_cata · 2026-11-19
- Study: Claude investigator agents correctly identify the purpose of most alignment evals — CFGeek · 2026-11-19
- 7,000+ human gaze trials used to train gaze-guided visual reasoner GGR — LakeBrenden · 2026-11-19