Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

FEVEROUS

Benchmarks

Task NameDataset NameSOTA ResultTrend
Fact-checkingFEVEROUS (test)
Macro F174.72
20
Fact-checkingFEVEROUS
F1 Macro89.4
14
Fact-checkingFEVEROUS-S
Macro F172.55
12
Claim VerificationFEVEROUS
Accuracy0.9567
12
Evidence RetrievalFEVEROUS 2021 (test)
Recall@2087.02
10
GenerationFEVEROUS n=250
3-Class Accuracy58.4
8
RetrievalFEVEROUS n=250
Success@538.8
8
Table Fact VerificationFEVEROUS (dev)
Accuracy82.4
8
Multi-hop Open-domain Question AnsweringFeverous
Accuracy0.88
6
Hint-sensitivity forecastingFEVEROUS 610 trajectories (held-out)
Spearman Correlation0.8
4
Claim VerificationFEVEROUS (test)
Accuracy83.8
4
Showing 11 of 11 rows