TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Models/ProphetNet

ProphetNet

Reported on 12 benchmarks across 3 tasks · 1 paper · 8 SOTA

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

Knowledge Base6 results

  • Text SummarizationonGigaWord
    ROUGE-1· uses extra data· 2020-01-13
    39.51
    best: 60.12 (OpenAI/o3-mini)
    SOTA
    ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingarXiv:2001.04063
  • Text SummarizationonCNN / Daily Mail
    ROUGE-1· uses extra data· 2020-01-13
    44.2
    best: 48.18 (Scrambled code + broken (alter))
    SOTA
    ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingarXiv:2001.04063
  • Text SummarizationonCNN / Daily Mail
    ROUGE-L· uses extra data· 2020-01-13
    41.3
    best: 45.35 (Scrambled code + broken (alter))
    SOTA
    ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingarXiv:2001.04063
  • Text SummarizationonGigaWord
    ROUGE-2· uses extra data· 2020-01-13
    20.42
    best: 54.22 (OpenAI/o3-mini)
    ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingarXiv:2001.04063
  • Text SummarizationonGigaWord
    ROUGE-L· uses extra data· 2020-01-13
    36.69
    best: 60.29 (Riple/Saanvi-v0.1)
    ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingarXiv:2001.04063
  • Text SummarizationonCNN / Daily Mail
    ROUGE-2· uses extra data· 2020-01-13
    21.17
    best: 24.02 (Pegasus)
    ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingarXiv:2001.04063

Natural Language Processing6 results

  • Abstractive Text SummarizationonCNN / Daily Mail
    ROUGE-1· uses extra data· 2020-01-13
    44.2
    best: 48.18 (Scrambled code + broken (alter))
    SOTA
    ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingarXiv:2001.04063
  • Abstractive Text SummarizationonCNN / Daily Mail
    ROUGE-L· uses extra data· 2020-01-13
    41.3
    best: 45.35 (Scrambled code + broken (alter))
    SOTA
    ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingarXiv:2001.04063
  • Question GenerationonSQuAD1.1
    BLEU-4· uses extra data· 2020-01-13
    23.91
    best: 25.41 (ERNIE-GENLARGE (beam size=5))
    SOTA
    ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingarXiv:2001.04063
  • Question GenerationonSQuAD1.1
    METEOR· uses extra data· 2020-01-13
    26.6
    best: 26.73 (BART (TextBox 2.0))
    SOTA
    ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingarXiv:2001.04063
  • Question GenerationonSQuAD1.1
    ROUGE-L· uses extra data· 2020-01-13
    52.3
    best: 52.8 (ProphetNet + ASGen)
    SOTA
    ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingarXiv:2001.04063
  • Abstractive Text SummarizationonCNN / Daily Mail
    ROUGE-2· uses extra data· 2020-01-13
    21.17
    best: 24.02 (Pegasus)
    ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingarXiv:2001.04063