FEVER
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
FEVER (held-out test)
39.37Accuracy
4
FEVER
91Macro F1
4
FEVER (val)
84.67Macro-F1
4
Fever
8,580ITC Score
4
Fever
88.3Recall@10
4
Fever
0.47Average Indexing Time (AIT)
4
FEVER (test)
99.7Accuracy
4
FEVER-HN
0.939MRR@10
4
FEVER
0.725AUROC
4
Fever Expert-Constructed Bayesian Network
32.89CMI
4
Fever
0Ratio Y/(N+Y)
4
FEVER
89.98Accuracy
3
FEVER group-balanced (evaluation)
44Bias Reduction (%)
3
FEVER
59.54Accuracy
3
FEVER
20.19QPS
3
FEVER-2 (test)
92.2Accuracy
3
FEVER-3 (test)
76.8Label Accuracy
3
FEVER Strict-Multi
72.1Label Accuracy (LA)
3
FEVER (dev)
93.55OFEVER
3
FEVER 2.0
73.3Accuracy
2
FEVER
46Evidence Swap Follow Accuracy
2
FEVER
59.93Accuracy
2