List your company

Uncertainty Estimation on MATH Camel

0.7354AUROCMATU
0.5153360.5724680.62960.686732Apr 9, 2026

Evaluation results

MethodLinks
MATUBase LLM=Llama3.1-8B2026.040.73540.7525
MATUBase LLM=Qwen2.5-7B2026.040.70890.9064
MATUBase LLM=GPT-4o2026.040.67970.616
Eigv(Agre)-WholeBase LLM=Qwen2.5-7B2026.040.67840.8963
P(true)Base LLM=Qwen2.5-7B2026.040.63510.8855
SAUP-MultipleBase LLM=Qwen2.5-7B2026.040.60780.8722
P(true)Base LLM=GPT-4o2026.040.58250.5592
Eigv(Agre)-finalBase LLM=GPT-4o2026.040.56980.5216
Eigv(Agre)-WholeBase LLM=GPT-4o2026.040.56320.5218
Eigv(Agre)-WholeBase LLM=Llama3.1-8B2026.040.56220.6346
SAUP-SingleBase LLM=Qwen2.5-7B2026.040.55970.8499
P(true)Base LLM=Llama3.1-8B2026.040.54210.6303
SAUP-MultipleBase LLM=Llama3.1-8B2026.040.52580.6427
SAUP-SingleBase LLM=Llama3.1-8B2026.040.52440.6374
Eigv(Agre)-finalBase LLM=Llama3.1-8B2026.040.52430.617
Eigv(Agre)-finalBase LLM=Qwen2.5-7B2026.040.52380.8466