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 6 · SOTA2 Research

Summarization

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

PubMed 2018 (test)

Oracle (ROUGE-2, F1)

55.05ROUGE-1

15

Feb 26, 2026

CNN/Daily Mail anonymized (test)

40.13ROUGE-1

15

Feb 26, 2026

OPEN-I and MIMIC-CXR combined (test)

56.5ROUGE-L

15

Feb 26, 2026

36.45ROUGE-L

14

Jun 2, 2026

Medical Random subset

25.04R-LCS

14

May 19, 2026

27.1ROUGE-2

14

Feb 27, 2026

17.4Token MAE

14

Feb 26, 2026

SCALENET (Layer-wise)

22.47ROUGE-Lsum

14

Feb 26, 2026

XLSum Korean (test)

Vanilla Decoding

20.7ROUGE-2

14

Jun 9, 2026

36.4ROUGE-2

14

Feb 27, 2026

BookSum Chapter Level

Extractive Oracle

42.68ROUGE-1

14

Feb 26, 2026

iEEG clinical dataset Background

46.3ROUGE-L

14

Feb 26, 2026

iEEG clinical dataset Foreground

60.9ROUGE-L

14

Feb 26, 2026

Visual Tasks (test)

75.66Text F1

13

Apr 16, 2026

46.4Score

13

Mar 30, 2026

40.35ROUGE-1

13

Feb 26, 2026

GovReport (test)

0.555ROUGE-1

13

Apr 28, 2026

ENTSUM 1.0 (test)

Lead3ent (Summary)

85.22ROUGE-1

13

Feb 26, 2026

BioPubSum (test)

39.2ROUGE-1

12

Jun 25, 2026

AIPubSum (test)

38.99ROUGE-1

12

Jun 25, 2026

CSPubSum (test)

39.9ROUGE-1

12

Jun 25, 2026

11895th Percentile Latency (ms)

12

Apr 23, 2026

47.5ROUGE-L

12

Apr 9, 2026

39.6ROUGE-L

12

Apr 9, 2026

19.67ROUGE-1

12

Feb 26, 2026

Page 6 of 19

Popular datasets

CNN/DM Xsum SamSum XLSum Pubmed

Follow for updates

Join SOTA2 Contact