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 13 · SOTA2 Research

Summarization

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

MobileLLM-Flash

16.89TLDR9+ Score

6

Jul 7, 2026

43.51ROUGE-1

6

Feb 26, 2026

SQUAD 1.1 (Memorization)

Chain of Summaries (CoS)

59F1 Score

6

Feb 26, 2026

TruthfulQA (Memorization)

Chain of Summaries (CoS)

46.06ROUGE-L

6

Feb 26, 2026

TriviaQA (Memorization)

Chain of Summaries (CoS)

52F1 Score (TriviaQA Mem)

6

Feb 26, 2026

21.6ROUGE-L

6

Jun 25, 2026

73.18Grade

6

Feb 26, 2026

SMOOTHIE-GLOBAL

14.5ROUGE-2

6

Feb 26, 2026

Extractive Blueprint

72.98ROUGE-L

6

Feb 26, 2026

12.56ROUGE-L

6

Feb 26, 2026

BookSum Average latest (test)

17.47Average ROUGE

6

Feb 26, 2026

BookSum Trun. latest (test)

16.68Avg ROUGE

6

Feb 26, 2026

BookSum No Trun. latest (test)

17.74Average ROUGE

6

Feb 26, 2026

44.33ROUGE-1

6

Feb 26, 2026

WikiLingua target languages

24.9AR Score

6

Feb 26, 2026

57.19R1

6

Feb 26, 2026

45.12ROUGE-1

6

Feb 26, 2026

Summarization Human Evaluation (test)

SelfGAN_Coop-MCTS

4Consistency

6

Feb 26, 2026

Gigaword sampled 200 instances (test)

2.846Informativeness

6

Feb 26, 2026

CL-SciSumm 2016 (test)

41.69ROUGE-2 Recall

6

Feb 26, 2026

Gigaword 100 randomly-selected articles (test)

3.55Informativity

6

Feb 26, 2026

CNN/DailyMail human evaluation (100 samples)

Soft Layer-Specific Multi-Task Summarization (MTL)

43Relevance Score

6

Feb 26, 2026

Gigaword two length limit

36.54ROUGE-1 F-Score

6

Feb 26, 2026

DUC 75-byte limit 2004

32.28ROUGE-1 Recall

6

Feb 26, 2026

92.8Win Rate

6

Feb 26, 2026

Page 13 of 19

Popular datasets

CNN/DM Xsum SamSum XLSum Pubmed

Follow for updates

Join SOTA2 Contact