NQ-Open
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
NQ-Open
0.93AUROC
141
NQ-Open (val)
49.62Accuracy
46
NQ-Open In-Domain (test)
58.13Precision
26
NQ-Open ID
57.34Precision
24
NQ-open Augmented (full-slice)
85.42Restate-hard
18
NQ-open v1.0 (test)
79.08A1
16
NQ-Open Out-of-distribution (test)
49.93Accuracy
15
NQ Open (test)
89.4AUROC
14
NQ-Open (out-of-domain)
0.705Precision
12
NQ-Open (test)
25.61Mean F1 Score
10
NQ-Open (eval)
50.43F1 Score
9
NQ-Open OOD (test)
82.81Exact Match (EM)
9
NQ-Open
17.5Exact Match (EM)
8
NQ-Open first 200 examples public (val)
6.8Mean Jaccard
2