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

20NG

Benchmarks

Task NameDataset NameSOTA ResultTrend
Topic Modeling20NG
NPMI0.24
33
Text Classification20NG
Accuracy89.5
20
Text Classification20ng (test)
Accuracy (Test)89.5
18
Class Incremental Learning20NG
Average Backward Transfer (BWT)29.77
13
Class Incremental Learning20NG (test)
Accuracy97.06
13
Topic Modeling20NG
IRBO99.6
13
Document Clustering20NG (test)
NMI0.437
13
Multi-class Calibration20NG (test)
NLL1.4
8
Document Clustering20NG
Top-Purity47.7
6
Out-of-Distribution Detection20NG (ID) vs IMDB (OOD)
AUROC99.78
4
Clustering20NG
ARI0.845
4
Clustering20NG-C
k-Means Loss951.96
2
Clustering20NG-B
K-Means Loss481.72
2
Clustering20NG-A
k-Means Loss193.46
2
Showing 14 of 14 rows