List your company

ResearchFormerly WizWand

UMWP

Benchmarks

TaskDatasetSOTA resultTrendResults

Hallucination Detection

UMWP

89.6ROC-AUC

28

Hallucination Detection (Math Word Problems)

UMWP

89.1F1 Score

12

Abstention in Question Answering

UMWP Underspecified Context

94Abstention F1

10

UMWP

0.62False Positive Rate

5