List your company

Language Modeling on Clustered Pre-training Data Cluster 11

22.002Perplexity (PPL)Split MoE
21.9898822.0716922.153522.23531Mar 19, 2026

Evaluation results

MethodLinks
Split MoEModel Variant=Split MoE2026.0322.002
BaseModel Variant=Base2026.0322.305