MAQA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
MAQA
0.04Hamming Distance
52
MAQA-ΔK−1
-0.149KL Divergence
48
MAQA ∆K−1
0.757KL Divergence AUC
28
MAQA
83.5Hamming AUC
28
Non-MAQA
77.53AUROC
20
MAQA {0, 1}^K
102.3Hamming Score
20
MAQA (test)
90.3BERTScore
12
MAQA
0.635Accuracy
7
MAQA
90.3BERTScore
6
MAQA* High-Ambiguity Subset H[p*] >= 1.5 (test)
65AUROC
5
MAQA (test)
63.5Accuracy
5