List your company

Question Answering on ASDiv

66.91AUROCTOKENSAR + CoT-UQ
46.442851.756457.0762.3836Feb 24, 2025

Evaluation results

MethodLinks
TOKENSAR + CoT-UQBackbone=Llama 3.1-8B, CoT-UQ=true2025.0266.9175.43
Probas-mean + CoT-UQBackbone=Llama 3.1-8B, CoT-UQ=true2025.0264.5273.95
TOKENSARBackbone=Llama 3.1-8B, CoT-UQ=false2025.0258.7168.6
Probas-meanBackbone=Llama 3.1-8B, CoT-UQ=false2025.0258.3469.03
P(True) + CoT-UQBackbone=Llama 3.1-8B, CoT-UQ=true2025.0253.264.15
P(True)Backbone=Llama 3.1-8B, CoT-UQ=false2025.0247.2360.28