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

Summarization research benchmarks – Page 15 · SOTA2 Research

Summarization

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Newsroom Biased

51.1ROUGE-L

5

Feb 26, 2026

CNN/DM Non-Biased

0.173ROUGE-L

5

Feb 26, 2026

27.1ROUGE-L

5

Feb 26, 2026

30Book Score

5

Feb 26, 2026

7.8Mul. News Score

5

Feb 26, 2026

0.876AUC-ROC

5

Feb 26, 2026

24.86Multilingual Score

5

Feb 26, 2026

WikiLingua GEM (test)

10.4ROUGE-2

5

Feb 26, 2026

4.93Fluency

5

Feb 26, 2026

55.61ROUGE-1

5

Feb 26, 2026

MEDIQA-ANS Pages (test)

State-of-the-art (Laskar et al.)

0.231ROUGE-1

5

Feb 26, 2026

MEDIQA-MAS (test)

32.9ROUGE-1

5

Feb 26, 2026

MEDIQA-QS (test)

State-of-the-art (Ben Abacha et al.)

35.14ROUGE-1

5

Feb 26, 2026

HealthCareMagic (test)

47.24ROUGE-1

5

Feb 26, 2026

State-of-the-art (Mrini et al.)

62.3ROUGE-1

5

Feb 26, 2026

Gigaword (evaluation)

40.4ROUGE-1

5

Feb 26, 2026

CNN non-anonymized (test)

30.4ROUGE-1

5

Feb 26, 2026

SuperSummary Things we Hide from the Light (Manual analysis)

Extract-Support

72.7Correct %

5

Feb 26, 2026

NSF-QA (far-field)

Dixtral (finetuned)

41.4ROUGE-L

4

Jun 17, 2026

0Response-level Language Confusion

4

Jun 9, 2026

0Response-level Language Confusion

4

Jun 9, 2026

0Response-level Language Confusion

4

Jun 9, 2026

0Response-level Language Confusion

4

Jun 9, 2026

0Response-level Language Confusion

4

Jun 9, 2026

0Response-level Language Confusion

4

Jun 9, 2026

Page 15 of 19

Popular datasets

CNN/DM Xsum SamSum XLSum Pubmed

Follow for updates

Join SOTA2 Contact