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/GPT-3 175B (Few-Shot)

GPT-3 175B (Few-Shot)

Reported on 13 benchmarks across 4 tasks · 1 paper · 8 SOTA

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

Natural Language Processing11 results

  • Machine TranslationonWMT2014 French-English
    BLEU· 2020-05-28
    39.2
    SOTA
    Language Models are Few-Shot LearnersarXiv:2005.14165
  • Machine TranslationonWMT2016 English-German
    BLEU· 2020-05-28
    29.7
    SOTA
    Language Models are Few-Shot LearnersarXiv:2005.14165
  • Machine TranslationonWMT2016 Romanian-English
    BLEU· 2020-05-28
    39.5
    SOTA
    Language Models are Few-Shot LearnersarXiv:2005.14165
  • Machine TranslationonWMT2016 German-English
    BLEU· 2020-05-28
    40.6
    SOTA
    Language Models are Few-Shot LearnersarXiv:2005.14165
  • Question AnsweringonStory Cloze
    Accuracy· 2020-05-28
    87.7
    best: 87.8 (Neo-6B (QA + WS))
    SOTA
    Language Models are Few-Shot LearnersarXiv:2005.14165
  • Question AnsweringonTriviaQA
    EM· uses extra data· 2020-05-28
    71.2
    best: 87.5 (Claude 2 (few-shot, k=5))
    SOTA
    Language Models are Few-Shot LearnersarXiv:2005.14165
  • Machine TranslationonWMT2014 English-French
    BLEU· 2020-05-28
    32.6
    best: 38.27 (BERT-fused NMT)
    Language Models are Few-Shot LearnersarXiv:2005.14165
  • Machine TranslationonWMT2016 English-Romanian
    BLEU· 2020-05-28
    21
    best: 35.2 (MASS (6-layer Transformer))
    Language Models are Few-Shot LearnersarXiv:2005.14165
  • Question AnsweringonMultiRC
    F1· 2020-05-28
    75.4
    best: 90.1 (PaLM 540B (finetuned) )
    Language Models are Few-Shot LearnersarXiv:2005.14165
  • Question AnsweringonRACE
    RACE-h· 2020-05-28
    46.8
    best: 69.6 (OCN_large)
    Language Models are Few-Shot LearnersarXiv:2005.14165
  • Natural Language InferenceonCommitmentBank
    Accuracy· 2020-05-28
    75.6
    best: 100 (PaLM 540B (finetuned))
    Language Models are Few-Shot LearnersarXiv:2005.14165

Medical2 results

  • Language ModellingonLAMBADA
    Accuracy· 2020-05-28
    86.4
    best: 89.7 (PaLM-540B (Few-Shot))
    SOTA
    Language Models are Few-Shot LearnersarXiv:2005.14165
  • Language ModellingonLAMBADA
    Perplexity· 2020-05-28
    1.92
    SOTA
    Language Models are Few-Shot LearnersarXiv:2005.14165