List your company

Uncertainty Estimation on MMLU Camel

0.7149AUROCMATU
0.497540.553970.61040.66683Apr 9, 2026

Evaluation results

MethodLinks
MATUBase LLM=Qwen2.5-7B2026.040.71490.8656
MATUBase LLM=Llama3.1-8B2026.040.70750.8427
P(true)Base LLM=Qwen2.5-7B2026.040.68460.8585
P(true)Base LLM=Llama3.1-8B2026.040.62070.7964
Eigv(Agre)-WholeBase LLM=Llama3.1-8B2026.040.5940.7646
SAUP-MultipleBase LLM=Qwen2.5-7B2026.040.56410.81
MATUBase LLM=GPT-4o2026.040.56040.3384
Eigv(Agre)-finalBase LLM=Qwen2.5-7B2026.040.55370.8023
SAUP-SingleBase LLM=Llama3.1-8B2026.040.54240.7361
Eigv(Agre)-WholeBase LLM=Qwen2.5-7B2026.040.5420.7995
Eigv(Agre)-finalBase LLM=GPT-4o2026.040.53650.3304
Eigv(Agre)-WholeBase LLM=GPT-4o2026.040.53410.3236
SAUP-MultipleBase LLM=Llama3.1-8B2026.040.52890.733
SAUP-SingleBase LLM=Qwen2.5-7B2026.040.52330.7749
Eigv(Agre)-finalBase LLM=Llama3.1-8B2026.040.51610.727
P(true)Base LLM=GPT-4o2026.040.50590.3183