NQ
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
NQ-Swap
84.7Exact Match (EM)
17
NQ-Open Open-Book
79.51Exact Match
17
NQ PIA (test)
53.2RACC
17
NQ-Open
5.13Delta NQ-Open
17
NQ
48.5Accuracy
17
NQ
75.96F1 Score
17
NQ
67Clean Accuracy (NQ)
16
NQ
100ATR
16
NQ
98ATR
16
NQ (IID)
100Recall
16
NQ (NIID-1)
99.75Recall
16
NQ
90.39ConR
16
NQ
86.21P@1
16
NQ
37.6QA-F1
16
NQ
41.88Exact Match (EM)
15
NQ
1.8ASR (PIA)
15
NQ poison @ Position 10, k=10 (test)
65.8Robustness Accuracy
15
NQ poison @ Position 1, k=10 (test)
64.2Robustness Accuracy
15
NQ
11RepeatScore
15
NQ
63.3F1 Score
14
NQ
5Target@5
14
NQ (test)
50.1EM
14
NQ (val)
62.1F1 Score
14
NQ (test)
15.818F1 Score
14
NQ (test)
8.339NDCG@5
14