Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

NaturalQA

Benchmarks

Task NameDataset NameSOTA ResultTrend
Question AnsweringNaturalQA
Accuracy38
66
Question AnsweringNaturalQA (test)
Accuracy67.4
39
Uncertainty EstimationNaturalQA
AUROC77
39
Question AnsweringNaturalQA
EM38.92
26
Abstention ClassificationNaturalQA (test)
Accuracy (Abs=0)100
9
Showing 5 of 5 rows