CNN Dailymail

Benchmarks

Task Name	Dataset Name	SOTA Result
Best feasible prompt identification	CNN/DailyMail (test)	Average Soft Constrained Reward0.172	72
Abstractive Summarization	CNN/DailyMail full length F-1 (test)	ROUGE-141.69	48
Open ended generation	CNN DailyMail	ROUGE-L24.3	40
Pareto prompt set identification	CNN/DailyMail	Hypervolume (HV)18.03	36
Language Generation	CNN/DailyMail	Accuracy27.16	35
Summarization	CNN/DailyMail (test)	ROUGE-L48	33
Uncertainty Quantification	CNN/DailyMail	Hamming AUC0.745	28
Summarization	CNN/DailyMail	Hamming Score-0.276	28
Abstractive Summarization	CNN/DailyMail	ROUGE-144.51	25
Summarization	CNN DailyMail	PRR0.45	22
Summarization	CNN/DailyMail	RougeL23.23	21
Abstractive Summarization	CNN/DailyMail Summarization	Hamming Distance1.597	20
Text Summarization	CNN/DailyMail	BA97.34	16
Reranking	CNN DailyMail	R-152.43	15
Context Compression	CNN/DailyMail	ROUGE-144.89	13
Text Summarization	CNN DailyMail	ROUGE-138.58	13
Long-context Modeling	CNN/DailyMail	Speedup3.85	12
Binary Classification (Assistive vs Creative)	CNN DailyMail	AUC99	12
Binary Classification (Human vs Creative)	CNN/DailyMail	AUC99	12
Binary Classification (Human vs Assistive)	CNN/DailyMail	AUC0.99	12
Attribution Quality Evaluation	CNN DailyMail	Log-Prob Drop1.371	12
Short-Generation	CNN/DailyMail	ROUGE-121.15	10
Length-Constrained Text Generation	CNN/DailyMail	Win Rate16.43	10
Text Generation	CNN/DailyMail (test)	LCTG Error Rate (E)3.18	10
Text Summarization	CNN/DailyMail (test)	ROUGE-133.23	9

Showing 25 of 52 rows