Knowledge
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
knowledge
99Accuracy
28
Knowledge Macro-aggregate
80.5Pass@1
22
knowledge 25% feature noise
98Accuracy
14
knowledge
98.2F1-score
14
Knowledge WikiDYK, PopQA, BLEND
37.03Accuracy
8
knowledge
20.2ARI
8
Knowledge
48.1Average Score
8
knowledge 0% noise (test)
98.3Accuracy
7
knowledge 0% noise
99.1F1-score
7
knowledge (test)
96.3Accuracy (test)
7
knowledge 25% feature noise
97.6F1 Score
7
knowledge 25% label noise
96F1 Score
7
KNOWLEDGE
88.7Accuracy@8
5