List your company

Language Modeling on Topic Domains (Held-out)

6.84Perplexity (PPL)Llama-8B (KOFF)
5.817212.721119.62526.5289May 27, 2026

Evaluation results

MethodLinks
Llama-8B (KOFF)Model Backbone=Llama-8B, Model Variant=KOFF, Global Sparsity Level=12%2026.056.84
Llama-8B (unpruned)Model Backbone=Llama-8B, Model Variant=unpruned, Global Sparsity Level=0%2026.057.93
Llama-3B (unpruned)Model Backbone=Llama-3B, Model Variant=unpruned, Global Sparsity Level=0%2026.058.74
Llama-3B (KOFF)Model Backbone=Llama-3B, Model Variant=KOFF, Global Sparsity Level=12%2026.058.85
Qwen-3B (KOFF)Model Backbone=Qwen-3B, Model Variant=KOFF, Global Sparsity Level=12%2026.0513.75
Qwen-3B (unpruned)Model Backbone=Qwen-3B, Model Variant=unpruned, Global Sparsity Level=0%2026.0514.14
Llama-8B (pruning only)Model Backbone=Llama-8B, Model Variant=pruning only, Global Sparsity Level=12%2026.0519.78
Llama-3B (pruning only)Model Backbone=Llama-3B, Model Variant=pruning only, Global Sparsity Level=12%2026.0524.01
Qwen-3B (pruning only)Model Backbone=Qwen-3B, Model Variant=pruning only, Global Sparsity Level=12%2026.0532.41