Abstractive Text Summarization on CNN / Daily Mail

Metric: ROUGE-2 (higher is better)

LeaderboardDataset

Loading chart...

Results

Hide extra data

Sort:

#	Model↕	ROUGE-2▼	Extra Data	Paper	Date↕	Code
1	Pegasus	24.02	No	Calibrating Sequence likelihood Improves Conditi...	2022-09-30	-
2	BRIO	23.55	No	BRIO: Bringing Order to Abstractive Summarization	2022-03-31	Code
3	SEASON	22.64	No	Salience Allocation as Guidance for Abstractive ...	2022-10-22	Code
4	PEGASUS + SummaReranker	22.61	No	SummaReranker: A Multi-Task Mixture-of-Experts R...	2022-03-13	Code
5	BART + SimCLS	22.15	No	SimCLS: A Simple Framework for Contrastive Learn...	2021-06-03	Code
6	BART + R-Drop	21.58	No	R-Drop: Regularized Dropout for Neural Networks	2021-06-28	Code
7	Fourier Transformer	21.55	No	Fourier Transformer: Fast Long Range Modeling by...	2023-05-24	Code
8	CoCoNet + CoCoPretrain	21.55	Yes	-	-	Code
9	T5	21.55	No	Exploring the Limits of Transfer Learning with a...	2019-10-23	Code
10	BART+R3F	21.53	No	Better Fine-Tuning by Reducing Representational ...	2020-08-06	Code
11	PEGASUS	21.47	Yes	PEGASUS: Pre-training with Extracted Gap-sentenc...	2019-12-18	Code
12	CoCoNet	21.41	No	-	-	Code
13	GLM-XXLarge	21.4	Yes	GLM: General Language Model Pretraining with Aut...	2021-03-18	Code
14	LongT5	21.4	No	LongT5: Efficient Text-To-Text Transformer for L...	2021-12-15	Code
15	ERNIE-GENLARGE (large-scale text corpora)	21.35	Yes	ERNIE-GEN: An Enhanced Multi-Flow Pre-training a...	2020-01-26	Code
16	BART	21.28	No	BART: Denoising Sequence-to-Sequence Pre-trainin...	2019-10-29	Code
17	MUPPET BART Large	21.25	No	Muppet: Massive Multi-task Representations with ...	2021-01-26	Code
18	ProphetNet	21.17	Yes	ProphetNet: Predicting Future N-gram for Sequenc...	2020-01-13	Code
19	ERNIE-GENLARGE	21.17	No	ERNIE-GEN: An Enhanced Multi-Flow Pre-training a...	2020-01-26	Code
20	PALM	21.12	No	PALM: Pre-training an Autoencoding&Autoregressiv...	2020-04-14	Code
21	UniLM	20.43	Yes	Unified Language Model Pre-training for Natural ...	2019-05-08	Code
22	UniLMv2	20.42	Yes	UniLMv2: Pseudo-Masked Language Models for Unifi...	2020-02-28	Code
23	Scrambled code + broken	20.39	No	Universal Evasion Attacks on Summarization Scoring	2022-10-25	Code
24	ERNIE-GENBASE	19.92	Yes	ERNIE-GEN: An Enhanced Multi-Flow Pre-training a...	2020-01-26	Code
25	Scrambled code + broken (alter)	19.84	No	Universal Evasion Attacks on Summarization Scoring	2022-10-25	Code
26	SRformer-BART	19.8	No	Segmented Recurrent Transformer: An Efficient Se...	2023-05-24	Code
27	BertSumExtAbs	19.6	Yes	Text Summarization with Pretrained Encoders	2019-08-22	Code
28	Two-Stage + RL	19.49	No	Pretraining-Based Natural Language Generation fo...	2019-02-25	Code
29	DCA	19.47	Yes	Deep Communicating Agents for Abstractive Summar...	2018-03-27	-
30	BERT-ext + abs + RL + rerank	19.08	No	Summary Level Training of Sentence Rewriting for...	2019-09-19	-
31	EditNet	19.03	No	An Editorial Network for Enhanced Document Summa...	2019-02-27	-
32	Selector & Pointer-Generator	18.74	Yes	Mixture Content Selection for Diverse Sequence G...	2019-09-04	Code
33	rnn-ext + RL	18.72	No	Fast Abstractive Summarization with Reinforce-Se...	2018-05-28	Code
34	Bottom-Up Summarization	18.68	No	Bottom-Up Abstractive Summarization	2018-08-31	Code
35	Subformer-base	18.3	No	-	-	-
36	Mask Attention Network	18.29	No	Mask Attention Networks: Rethinking and Strength...	2021-03-25	Code
37	Li et al.	18.18	No	-	-	-
38	Li et al.	18.02	No	-	-	-
39	ROUGESal+Ent RL	18	No	Multi-Reward Reinforced Summarization with Salie...	2018-04-17	-
40	end2end w/ inconsistency loss	17.97	No	A Unified Model for Extractive and Abstractive S...	2018-05-16	Code
41	RL + pg + cbdec	17.87	No	Closed-Book Training to Improve Summarization En...	2018-09-12	-
42	rnn-ext + abs + RL + rerank	17.8	No	Fast Abstractive Summarization with Reinforce-Se...	2018-05-28	Code
43	Pointer + Coverage + EntailmentGen + QuestionGen	17.64	No	Soft Layer-Specific Multi-Task Summarization wit...	2018-05-28	-
44	LEAD-3	17.62	No	Abstractive Text Summarization Using Sequence-to...	2016-02-19	Code
45	ML+RL ROUGE+Novel, with LM	17.38	No	Improving Abstraction in Text Summarization	2018-08-23	-
46	PTGEN + Coverage	17.28	No	Get To The Point: Summarization with Pointer-Gen...	2017-04-14	Code
47	PTGEN + Coverage	17.28	No	Get To The Point: Summarization with Pointer-Gen...	2017-04-14	Code
48	Pointer-Generator + Coverage	17.28	No	Get To The Point: Summarization with Pointer-Gen...	2017-04-14	Code
49	Dynamic Conv	16.25	No	Pay Less Attention with Lightweight and Dynamic ...	2019-01-29	Code
50	Transformer	16.06	No	Attention Is All You Need	2017-06-12	Code

#1PegasusSOTA
24.02
ROUGE-2· 2022-09-30
Calibrating Sequence likelihood Improves Conditional Language Generation
#2BRIOSOTA
23.55
ROUGE-2· 2022-03-31
BRIO: Bringing Order to Abstractive Summarization Code
#3SEASON
22.64
ROUGE-2· 2022-10-22
Salience Allocation as Guidance for Abstractive Summarization Code
#4PEGASUS + SummaRerankerSOTA
22.61
ROUGE-2· 2022-03-13
SummaReranker: A Multi-Task Mixture-of-Experts Re-ranking Framework for Abstractive Summarization Code
#5BART + SimCLSSOTA
22.15
ROUGE-2· 2021-06-03
SimCLS: A Simple Framework for Contrastive Learning of Abstractive Summarization Code
#6BART + R-Drop
21.58
ROUGE-2· 2021-06-28
R-Drop: Regularized Dropout for Neural Networks Code
#7Fourier Transformer
21.55
ROUGE-2· 2023-05-24
Fourier Transformer: Fast Long Range Modeling by Removing Sequence Redundancy with FFT Operator Code
#8CoCoNet + CoCoPretrain
21.55
ROUGE-2· Extra Data
No paperCode
#9T5SOTA
21.55
ROUGE-2· 2019-10-23
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer Code
#10BART+R3F
21.53
ROUGE-2· 2020-08-06
Better Fine-Tuning by Reducing Representational Collapse Code
#11PEGASUS
21.47
ROUGE-2· Extra Data· 2019-12-18
PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization Code
#12CoCoNet
21.41
ROUGE-2
No paperCode
#13GLM-XXLarge
21.4
ROUGE-2· Extra Data· 2021-03-18
GLM: General Language Model Pretraining with Autoregressive Blank Infilling Code
#14LongT5
21.4
ROUGE-2· 2021-12-15
LongT5: Efficient Text-To-Text Transformer for Long Sequences Code
#15ERNIE-GENLARGE (large-scale text corpora)
21.35
ROUGE-2· Extra Data· 2020-01-26
ERNIE-GEN: An Enhanced Multi-Flow Pre-training and Fine-tuning Framework for Natural Language Generation Code
#16BART
21.28
ROUGE-2· 2019-10-29
BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension Code
#17MUPPET BART Large
21.25
ROUGE-2· 2021-01-26
Muppet: Massive Multi-task Representations with Pre-Finetuning Code
#18ProphetNet
21.17
ROUGE-2· Extra Data· 2020-01-13
ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training Code
#19ERNIE-GENLARGE
21.17
ROUGE-2· 2020-01-26
ERNIE-GEN: An Enhanced Multi-Flow Pre-training and Fine-tuning Framework for Natural Language Generation Code
#20PALM
21.12
ROUGE-2· 2020-04-14
PALM: Pre-training an Autoencoding&Autoregressive Language Model for Context-conditioned Generation Code
#21UniLMSOTA
20.43
ROUGE-2· Extra Data· 2019-05-08
Unified Language Model Pre-training for Natural Language Understanding and Generation Code
#22UniLMv2
20.42
ROUGE-2· Extra Data· 2020-02-28
UniLMv2: Pseudo-Masked Language Models for Unified Language Model Pre-Training Code
#23Scrambled code + broken
20.39
ROUGE-2· 2022-10-25
Universal Evasion Attacks on Summarization Scoring Code
#24ERNIE-GENBASE
19.92
ROUGE-2· Extra Data· 2020-01-26
ERNIE-GEN: An Enhanced Multi-Flow Pre-training and Fine-tuning Framework for Natural Language Generation Code
#25Scrambled code + broken (alter)
19.84
ROUGE-2· 2022-10-25
Universal Evasion Attacks on Summarization Scoring Code
#26SRformer-BART
19.8
ROUGE-2· 2023-05-24
Segmented Recurrent Transformer: An Efficient Sequence-to-Sequence Model Code
#27BertSumExtAbs
19.6
ROUGE-2· Extra Data· 2019-08-22
Text Summarization with Pretrained Encoders Code
#28Two-Stage + RLSOTA
19.49
ROUGE-2· 2019-02-25
Pretraining-Based Natural Language Generation for Text Summarization Code
#29DCASOTA
19.47
ROUGE-2· Extra Data· 2018-03-27
Deep Communicating Agents for Abstractive Summarization
#30BERT-ext + abs + RL + rerank
19.08
ROUGE-2· 2019-09-19
Summary Level Training of Sentence Rewriting for Abstractive Summarization
#31EditNet
19.03
ROUGE-2· 2019-02-27
An Editorial Network for Enhanced Document Summarization
#32Selector & Pointer-Generator
18.74
ROUGE-2· Extra Data· 2019-09-04
Mixture Content Selection for Diverse Sequence Generation Code
#33rnn-ext + RL
18.72
ROUGE-2· 2018-05-28
Fast Abstractive Summarization with Reinforce-Selected Sentence Rewriting Code
#34Bottom-Up Summarization
18.68
ROUGE-2· 2018-08-31
Bottom-Up Abstractive Summarization Code
#35Subformer-base
18.3
ROUGE-2
No paper
#36Mask Attention Network
18.29
ROUGE-2· 2021-03-25
Mask Attention Networks: Rethinking and Strengthen Transformer Code
#37Li et al.
18.18
ROUGE-2
No paper
#38Li et al.
18.02
ROUGE-2
No paper
#39ROUGESal+Ent RL
18
ROUGE-2· 2018-04-17
Multi-Reward Reinforced Summarization with Saliency and Entailment
#40end2end w/ inconsistency loss
17.97
ROUGE-2· 2018-05-16
A Unified Model for Extractive and Abstractive Summarization using Inconsistency Loss Code
#41RL + pg + cbdec
17.87
ROUGE-2· 2018-09-12
Closed-Book Training to Improve Summarization Encoder Memory
#42rnn-ext + abs + RL + rerank
17.8
ROUGE-2· 2018-05-28
Fast Abstractive Summarization with Reinforce-Selected Sentence Rewriting Code
#43Pointer + Coverage + EntailmentGen + QuestionGen
17.64
ROUGE-2· 2018-05-28
Soft Layer-Specific Multi-Task Summarization with Entailment and Question Generation
#44LEAD-3SOTA
17.62
ROUGE-2· 2016-02-19
Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyond Code
#45ML+RL ROUGE+Novel, with LM
17.38
ROUGE-2· 2018-08-23
Improving Abstraction in Text Summarization
#46PTGEN + Coverage
17.28
ROUGE-2· 2017-04-14
Get To The Point: Summarization with Pointer-Generator Networks Code
#47PTGEN + Coverage
17.28
ROUGE-2· 2017-04-14
Get To The Point: Summarization with Pointer-Generator Networks Code
#48Pointer-Generator + Coverage
17.28
ROUGE-2· 2017-04-14
Get To The Point: Summarization with Pointer-Generator Networks Code
#49Dynamic Conv
16.25
ROUGE-2· 2019-01-29
Pay Less Attention with Lightweight and Dynamic Convolutions Code
#50Transformer
16.06
ROUGE-2· 2017-06-12
Attention Is All You Need Code