NQ
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
NQ
35.18Accuracy
6
NQ
100Answerability Rate
6
NQ Benign (test)
71.6Accuracy
6
NQ
12.82Score
6
NQ
52.2Hit Rate
6
NQ
25.5F1 (token-overlap)
6
NQ
65PIA RAcc
6
NQ
49.1ASR
6
NQ
67.6Reverse Score
6
NQ
71.22HIT@1
6
NQ (val)
35.8EM
6
NQ passage retrieval setting from DPR
86.4Recall@20
6
NQ
66.8Acc@5
6
NQ
75.8MRR
6
NQ
31.1Accuracy (Acc)
5
NQ in-domain (test)
44Accuracy
5
NQ
10Exact Match (EM)
5
NQ
95F1 Score
5
NQ
0.769E-SC
5
NQ
0.358E-SC
5
NQ Non-IID
76.33Accuracy
5
NQ IID
74.8Accuracy
5
NQ 3.5K (test)
0.261EM
5
NQ simplified
6Naïve RL Score
5
NQ
57.8EM
5