List your company

General Knowledge Reasoning on MMLU-Pro (test)

37.72AccuracyIntroLLM
14.600820.602926.60532.6071Feb 13, 2026

Evaluation results

MethodLinks
IntroLLMBackbone=Qwen3-4B-Base2026.0237.72
TampoBackbone=Qwen3-4B-Base, temperature range=0.6-1.52026.0237.23
GRPOBackbone=Qwen3-4B-Base, temperature (τ)=1.02026.0236.6
EADBackbone=Qwen3-4B-Base, annealing range=1.2→0.12026.0236.22
IntroLLMBackbone=Qwen3-1.7B-Base2026.0228.89
EADBackbone=Qwen3-1.7B-Base, annealing range=1.2→0.12026.0228.66
GRPOBackbone=Qwen3-1.7B-Base, temperature (τ)=1.02026.0228.39
TampoBackbone=Qwen3-1.7B-Base, temperature range=0.6-1.52026.0227.16
BaseBackbone=Qwen3-4B-Base2026.0225.06
BaseBackbone=Qwen3-1.7B-Base2026.0215.49