MASSIVE
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
MASSIVE
80.36Micro F1
64
Massive (test)
78.88NMI
20
MASSIVE (test)
78.6Accuracy
18
Massive-D
78Accuracy
17
MASSIVE (test)
89.47In-Scope Accuracy
17
MASSIVE Slotfill
57.3F1
14
MassiveIntentClassification
77.08Accuracy
11
MASSIVE Intent
0.62ARI
10
MASSIVE (unsupervised)
79.15Accuracy
9
MASSIVE (test)
100Guaranteed Test Coverage (alpha=0.10)
8
MASSIVE-Intent (test)
80.73CFT Score
8
MASSIVE-Slot (test)
62.54CFT
8
MASSIVE Intent
80.7Accuracy
8
MASSIVE
66In-Scope Accuracy
8
Massive I
60.5Accuracy
7
MASSIVE W5H2
0Cost/1K
7
Massive (I)
0.781NMI
6
Massive
68Label Quality
5
MASSIVE
87.6F1-Macro
5
MASSIVE (full)
87.6F1-Macro
5
MASSIVE W5H2 (test)
97.3Accuracy
4
Massive (test)
0.968AUROC
4
MASSIVE (test)
0.059ECE
4
MASSIVE
0.586ECE (Wrong Samples)
4
MASSIVE (test)
0.3ARI
4