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

Language Modeling research benchmarks – Page 37 · SOTA2 Research

Language Modeling

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

GigaWord v5 (test)

Relational Memory Core

38.3Perplexity

2

Feb 26, 2026

Project Gutenberg (test)

Relational Memory Core

42Perplexity

2

Feb 26, 2026

Project Gutenberg (val)

Relational Memory Core

39.2Perplexity

2

Feb 26, 2026

One Billion Word Benchmark (train)

36.39Perplexity

2

Feb 26, 2026

18.54Perplexity

1

Jun 16, 2026

2.24Base CE

1

Feb 26, 2026

3.48Base CE

1

Feb 26, 2026

29.37Perplexity

1

Feb 26, 2026

83.41PPL

1

Feb 26, 2026

154.81PPL

1

Feb 26, 2026

52.23PPL

1

Feb 26, 2026

C++ code (test)

0.605BPC (df)

1

Feb 26, 2026

Python code (test)

0.51BPC (df)

1

Feb 26, 2026

Flores-200 Urdu (devtest)

0.879BPC df

1

Feb 26, 2026

Flores-200 Basque (devtest)

1.243BPC (df)

1

Feb 26, 2026

Flores-200 Indonesian (devtest)

0.981BPC (df)

1

Feb 26, 2026

Flores-200 Vietnamese (devtest)

0.794BPC (df)

1

Feb 26, 2026

Flores-200 Spanish (devtest)

0.881BPC (df)

1

Feb 26, 2026

Flores-200 French (devtest)

0.8BPC (df)

1

Feb 26, 2026

Flores-200 Chinese (devtest)

1.208BPC (df)

1

Feb 26, 2026

Wikipedia articles En (test)

1.108BPC (df)

1

Feb 26, 2026

Tiny-Shakespeare

—

—Perplexity

0

Feb 26, 2026

Language Modeling Benchmark Suite (test)

—

—DM Mathematics Score

0

Feb 26, 2026

Page 37 of 37

Popular datasets

WikiText-103 WikiText-2 C4 LAMBADA WikiText

Follow for updates

Join SOTA2 Contact