NQ
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
NQ
46.9EM
54
NQ320k
48.92Hits@1
54
NQ
54.8Exact Match (EM)
52
NQ
54.33Reliability Score (RS)
52
NQ-Open
47.4Exact Match (EM)
48
NQ-DomainLoRA PMD-BENCH
89.7Accuracy
45
NQ
89Clean Accuracy
45
NQ
40.06F1 Score
45
NQ
11.76CRR
42
NQ
67.8F1
42
NQ (test)
44.38EM
39
NQ-Swap
73Accuracy
38
NQ
49.1Accuracy
37
NQ
80.8Accuracy
37
NQ Swap
89.63CF Score
36
NQ-Open
48.34Accuracy
36
NQ* (test)
40.09RAG F1
36
NQ
90.6NQ Recall (%)
36
NQ Poison Attack (test)
50.4Attack Success Rate
35
NQ
763Correlation Score
35
NQ-Open
83AUROC
32
NQ
3.51Speedup
32
NQ-open
75.88AUROC
32
NQ (test)
70.2Accuracy
30
NQ
99Precision
30