Codesota · Natural Language Processing · Text Summarization · CNN/DailyMailTasks/Natural Language Processing/Text Summarization
Text Summarization · benchmark dataset · 2015 · EN

CNN/DailyMail Summarization.

300K news articles with multi-sentence summaries. Standard benchmark for abstractive summarization.

Saturated benchmark· last significant update Jun 2022

ROUGE-based evaluation is saturated. No significant improvements since 2022. Modern summarization uses LLM-as-judge (G-Eval), human preference evaluations, or factual consistency metrics.

Paper ↗Download datasetSubmit a result ↵
§ 01 · Leaderboard

Best published scores.

15 results indexed across 3 metrics. Shaded row marks current SOTA; ties broken by submission date.

Primary
rouge-1 · higher is better
All metrics
rouge-1, rouge-2, rouge-l
rouge-1· primary
6 rows
#ModelOrgSubmittedPaper / coderouge-1
01BRIOOSSYale NLPMar 2022BRIO: Bringing Order to Abstractive Summarization47.78
02GPT-4oAPIOpenAIMar 2023GPT-4 Technical Report46.30
03Gemini 1.5 ProAPIGoogleFeb 2024Gemini 1.5: Unlocking multimodal understanding across mi…45.80
04Llama 3.1 405BOSSMetaJul 2024The Llama 3 Herd of Models45.10
05Qwen2 72BAlibabaJul 2024Qwen2 Technical Report44.70
06PEGASUS-LargeOSSGoogleDec 2019PEGASUS: Pre-training with Extracted Gap-sentences for A…44.17
rouge-2
3 rows
#ModelOrgSubmittedPaper / coderouge-2
01BRIOOSSYale NLPMar 2022BRIO: Bringing Order to Abstractive Summarization23.55
02GPT-4oAPIOpenAIMar 2023GPT-4 Technical Report22.10
03PEGASUS-LargeOSSGoogleDec 2019PEGASUS: Pre-training with Extracted Gap-sentences for A…21.47
rouge-l
6 rows
#ModelOrgSubmittedPaper / coderouge-l
01BRIOOSSYale NLPMar 2022BRIO: Bringing Order to Abstractive Summarization44.57
02GPT-4oAPIOpenAIMar 2023GPT-4 Technical Report43.40
03Gemini 1.5 ProAPIGoogleFeb 2024Gemini 1.5: Unlocking multimodal understanding across mi…43
04Llama 3.1 405BOSSMetaJul 2024The Llama 3 Herd of Models42.30
05Qwen2 72BAlibabaJul 2024Qwen2 Technical Report41.80
06PEGASUS-LargeOSSGoogleDec 2019PEGASUS: Pre-training with Extracted Gap-sentences for A…41.11
Fig 2 · Rows sorted by score within each metric. Shaded row marks SOTA. Dates reflect model or paper release where available, otherwise the date Codesota accessed the source.
§ 03 · Progress

2 steps
of state of the art.

Each row below marks a model that broke the previous record on rouge-1. Intermediate submissions are kept in the leaderboard above; only SOTA-setting entries are re-listed here.

Higher scores win. Each subsequent entry improved upon the previous best.

SOTA line · rouge-1
  1. Dec 18, 2019PEGASUS-LargeGoogle44.17
  2. Mar 31, 2022BRIOYale NLP47.78
Fig 3 · SOTA-setting models only. 2 entries span Dec 2019 → Mar 2022.
§ 04 · Literature

6 papers
tied to this benchmark.

Every paper below corresponds to at least one row in the leaderboard above. Click through for the arXiv preprint and, when available, the reference implementation.

§ 06 · Contribute

Have a score that beats
this table?

Submit a checkpoint and a reproduction script. We will run it, publish the score, and — if it takes the top — annotate the step on the progress chart with your name.

Submit a result ↵Read submission guide
What a submission needs
  • 01A public checkpoint or API endpoint
  • 02A reproduction script with frozen commit + seed
  • 03Declared evaluation environment (Python, deps)
  • 04One row per metric declared by this dataset
  • 05A contact so we can follow up on discrepancies
CNN/DailyMail — Text Summarization benchmark · Codesota | CodeSOTA