List your company

Language Modeling on Clustered Pre-training Data Cluster 13

19.461Perplexity (PPL)Split MoE
19.4576819.4800919.502519.52491Mar 19, 2026

Evaluation results

MethodLinks
Split MoEModel Variant=Split MoE2026.0319.461
BaseModel Variant=Base2026.0319.544