DANCER LSTM

Reported on 6 benchmarks across 1 task · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Knowledge Base6 results

Text SummarizationonArxiv HEP-TH citation graph
ROUGE-1· 2020-04-13
41.87
best: 50.95 (Top Down Transformer (AdaPool) (464M))
A Divide-and-Conquer Approach to the Summarization of Long Documents arXiv:2004.06190
Text SummarizationonArxiv HEP-TH citation graph
ROUGE-2· 2020-04-13
15.92
best: 21.93 (Top Down Transformer (AdaPool) (464M))
A Divide-and-Conquer Approach to the Summarization of Long Documents arXiv:2004.06190
Text SummarizationonArxiv HEP-TH citation graph
ROUGE-L· 2020-04-13
37.61
best: 45.61 (Top Down Transformer (AdaPool) (464M))
A Divide-and-Conquer Approach to the Summarization of Long Documents arXiv:2004.06190
Text SummarizationonPubmed
ROUGE-1· 2020-04-13
44.09
best: 51.05 (Top Down Transformer (AdaPool) (464M))
A Divide-and-Conquer Approach to the Summarization of Long Documents arXiv:2004.06190
Text SummarizationonPubmed
ROUGE-2· 2020-04-13
17.69
best: 24.76 (LongT5)
A Divide-and-Conquer Approach to the Summarization of Long Documents arXiv:2004.06190
Text SummarizationonPubmed
ROUGE-L· 2020-04-13
40.27
best: 46.67 (LongT5)
A Divide-and-Conquer Approach to the Summarization of Long Documents arXiv:2004.06190