List your company

Uncertainty Quantification on MMLU OOD via Math Prompts

67.7AUROCMATU
58.96461.23263.565.768Apr 9, 2026

Evaluation results

MethodLinks
MATUBase Model=Qwen2.5-7B, Protocol=strict distribution shift2026.0467.7
Eigv(Agre)-finalBase Model=Qwen2.5-7B, Protocol=strict distribution shift2026.0463.55
SAUP-MultipleBase Model=Qwen2.5-7B, Protocol=strict distribution shift2026.0462.01
SAUP-SingleBase Model=Qwen2.5-7B, Protocol=strict distribution shift2026.0459.3