Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Natural Questions

Benchmarks

Task NameDataset NameSOTA ResultTrend
Open Question AnsweringNatural Questions (NQ) (test)
Exact Match (EM)58.4
134
Over-refusal EvaluationNQ (Natural Questions)
ORR0
72
Question AnsweringNatural Questions (test)
EM61.65
72
Retrieval Attack DefenseNatural Questions (NQ)
ASR0
64
RetrievalNatural Questions (test)
Top-5 Recall92.1
62
Question AnsweringNQ (Natural Questions) (test)
Accuracy68.6
60
Question AnsweringNQ (Natural Questions)
EM78.3
55
Question AnsweringNatural Questions
EM70.58
52
Open Domain Question AnsweringNatural Questions (NQ)
Exact Match (EM)51.4
46
Question AnsweringNatural Questions (NQ) (test)
Robust Accuracy68
45
Passage retrievalNatural Questions (NQ) (test)
Top-20 Accuracy85.2
45
Embedding AlignmentNatural Questions (test)
Top-1 Accuracy100
40
Question AnsweringNatural Questions (NQ)
Accuracy49.3
36
Open-QA EvaluationEVOUNA-NaturalQuestions
F1 Score97.9
35
Question AnsweringNatural Questions (NQ) (test)
Exact Match60.4
35
Open-Domain Question AnsweringNQ (Natural Questions)
EM51.4
33
Question AnsweringNQ (Natural Questions)
EM42.5
28
Passage RetrievalNatural Questions (NQ)
Top-10 Accuracy66.59
28
Closed-book Question AnsweringNatural Questions (test)
Accuracy29.9
27
Information RetrievalNatural Questions (test)
Recall@2086.1
25
Single-hop QANQ (Natural Questions)
EM72
22
Knowledge EvaluationNatural Questions (NQ) (Evaluation)
Accuracy59.4
22
Extractive Question AnsweringNatural Questions MRQA
F1 Score81
22
Question AnsweringNatural Questions
Accuracy44.6
21
RAG Question AnsweringNQ (Natural Questions)
F1 Score54.06
20
Showing 25 of 84 rows