Golden MoE: Optimal Expert Selection at the Natural Equilibrium I=1/e
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Router learns to concentrate in zone [0.2123, 0.5], observed 36.8% = 1/e. CIFAR +4.8% vs Top-K. Overtakes Top-K at E=32 (13% faster), gap widens at E=64+.
Fields of Study & Subjects
Advanced Neural Network Applications
Domain Adaptation and Few-Shot Learning
Machine Learning and Data Classification
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 27 March 2026
- Language
- EN
- Document type
- Preprint