List your company

Language Modeling on Clustered Pre-training Data Cluster 5

19.55Perplexity (PPL)Split MoE
19.5464419.5704719.594519.61853Mar 19, 2026

Evaluation results

MethodLinks
Split MoEModel Variant=Split MoE2026.0319.55
BaseModel Variant=Base2026.0319.639