A new MoE training method integrates expert-level losses and partial online updates to improve forecasting accuracy and efficiency over standard statistical and neural models.
Switch transformers: Scaling to trillion parameter models with simple and efficient sparsity.Journal of Machine Learning Research, 23 (120):1–39
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Fast Training of Mixture-of-Experts for Time Series Forecasting via Expert Loss Integration
A new MoE training method integrates expert-level losses and partial online updates to improve forecasting accuracy and efficiency over standard statistical and neural models.