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

AG News

Benchmarks

Task NameDataset NameSOTA ResultTrend
Text ClassificationAG-NEWS
Accuracy94.1
248
Text ClassificationAG News (test)
Accuracy94.7
228
Topic ClassificationAG-News
Accuracy95.58
225
Topic ClassificationAG News (test)
Accuracy94.91
98
Text ClassificationAG NEWS RoBERTa-large (test)
CACC95.6
44
Membership Inference AttackAG News (test)
AUC0.909
43
Counterfactual GenerationAG News
LFR0.915
37
Language ModelingAG News
PPL4.76
36
News ClassificationAG News (test)
Accuracy91.7
34
Multi-class text classificationAG News
Micro-F10.917
33
Counterfactual GenerationAG News (test)
SLFR98
29
Text Adversarial Example DetectionAg-News
TPR@10100
28
Language ModelingAG News (val)
Perplexity18.19
28
Text Anomaly DetectionNLPAD-AGNews
AUROC94.84
25
Faithfulness EvaluationAG-news (test)
Rate of Label Changes2
24
Adversarial Text DetectionAG News
F1 Score96.7
24
Concept LearningAG News
Training Time234
21
Text embeddingAG News
t-value54.63
20
Text ClassificationAG News 40 labels
Top-1 Error Rate0.1067
19
Adversarial DetectionAG News
F1 Score95.7
18
Short Text ClusteringAG News (test)
Accuracy86.53
18
Class-Conditional Language GenerationAG News
MAUVE (World)0.963
16
Real-time latency evaluationAG-News
Latency (s)7
15
Topic ClassificationAG News (test)
Badnets CACC93.79
15
Language ModelingAG News (test)
Perplexity52.09
14
Showing 25 of 72 rows