Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Hate speech classification research benchmarks · SOTA2 Research

Hate speech classification

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

CM, EN, and HI (test)

Hing-RoBERTa Trilingual

0.693Consistency

42

Mar 23, 2026

Hindi (HI) (test)

Hing-RoBERTa Mixed

71.96Macro F1

42

Mar 23, 2026

English (EN) (test)

XLM-R Trilingual

70.03Macro F1 Score

42

Mar 23, 2026

Code-mixed (CM) (test)

XLM-R Trilingual

69.72Macro F1

42

Mar 23, 2026

99AUC

24

May 4, 2026

CREDENCE-DeBERTa

0.763Accuracy

21

Apr 28, 2026

Stormfront (test)

0.921AUC

16

Feb 26, 2026

Measuring Hate Speech (MHS) (test)

0.819AUC

16

Feb 26, 2026

HateCheck (test)

95.6AUC

16

Feb 26, 2026

0.912AUC

16

Feb 26, 2026

MHS small models

61.99F1 Score

14

May 27, 2026

70.6F1 Score

14

May 27, 2026

Implicit Hate (test)

0.67Macro F1

13

Apr 23, 2026

Latent Hate (test)

SMARTER (Llama_DPO-256)

69Macro F1 Score

13

Apr 23, 2026

HateXplain (test)

72Macro-F1

13

Apr 23, 2026

79.71F1 Score

12

Feb 26, 2026

97Precision

8

Feb 26, 2026

Socio Multi-Hot

75.9Precision

6

Apr 21, 2026

Ethos Benchmark

91Precision

5

Feb 26, 2026

HateXplain standard (test)

68.6Accuracy

5

Feb 26, 2026

0.938AUC

4

Feb 26, 2026

Popular datasets

ETHOS HateXplain MHS ToxiGen English (EN)

Follow for updates

Join SOTA2 Contact