List your company

Language Understanding on MMLU (NLL)

0.94NLLPPT (Distribution) + TS
0.89921.17461.451.7254May 26, 2026

Evaluation results

MethodLinks
PPT (Distribution) + TSBackbone=QWEN2-7B, Temperature Scaling=true2026.050.94
PPT (Distribution) + TSBackbone=LLAMA3-8B, Temperature Scaling=true2026.050.96
PPT (Distribution)Backbone=LLAMA3-8B, Temperature Scaling=false2026.051.01
Base + TSBackbone=LLAMA3-8B, Temperature Scaling=true2026.051.01
Base + TSBackbone=QWEN2-7B, Temperature Scaling=true2026.051.01
PPT (Distribution)Backbone=QWEN2-7B, Temperature Scaling=false2026.051.12
PPT (Mean)Backbone=LLAMA3-8B, Temperature Scaling=false2026.051.27
PPT (Mean)Backbone=QWEN2-7B, Temperature Scaling=false2026.051.67
BaseBackbone=LLAMA3-8B, Temperature Scaling=false2026.051.69
BaseBackbone=QWEN2-7B, Temperature Scaling=false2026.051.96