Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Multi-News

Benchmarks

Task NameDataset NameSOTA ResultTrend
Abstractive SummarizationMulti-News
ROUGE-221.1
47
Multi-document SummarizationMulti-News (test)
ROUGE-221.7
45
Indirect Prompt InjectionMulti-News
ASR100
42
Long-context language generationMulti-News
Average Acceptance Length (τ)3.51
25
Summarization FaithfulnessMulti-News
SummaC38.02
12
Faithfulness EvaluationMulti-News (test)
SummaC38.5
12
Multi-Document SummarizationMulti-News 256 (test)
ROUGE-146
12
Abstractive SummarizationMulti-News 56k samples (test)
ROUGE Score20.72
12
News SummarizationMulti-News
ROUGE-147.52
10
Extractive SummarizationMulti-News (test)
ROUGE-149.9
9
Topic GenerationMulti-News
Average Aggregate Score0.524
8
Multi-document summarizationMulti-News (test)
Informativeness150
4
Showing 12 of 12 rows