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/SpanBERT

SpanBERT

Reported on 25 benchmarks across 10 tasks · 4 papers · 11 SOTA

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

Natural Language Processing25 results

  • Coreference ResolutiononOntoGUM
    Avg F1· 2021-06-02
    64.6
    best: 68.2 (MTL-coref)
    SOTA
    OntoGUM: Evaluating Contextualized SOTA Coreference Resolution on 12 More GenresarXiv:2106.00933
  • Recognizing Emotion Cause in ConversationsonRECCON
    Exact Span F1· 2020-12-22
    34.64
    SOTA
    Recognizing Emotion Cause in ConversationsarXiv:2012.11820
  • Recognizing Emotion Cause in ConversationsonRECCON
    F1· 2020-12-22
    75.71
    SOTA
    Recognizing Emotion Cause in ConversationsarXiv:2012.11820
  • Recognizing Emotion Cause in ConversationsonRECCON
    F1(Neg)· 2020-12-22
    86.02
    SOTA
    Recognizing Emotion Cause in ConversationsarXiv:2012.11820
  • Recognizing Emotion Cause in ConversationsonRECCON
    F1(Pos)· 2020-12-22
    60
    SOTA
    Recognizing Emotion Cause in ConversationsarXiv:2012.11820
  • Relation ExtractiononRe-TACRED
    F1· 2019-07-24
    85.3
    best: 91.4 (EXOBRAIN)
    SOTA
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Question AnsweringonNaturalQA
    F1· 2019-07-24
    82.5
    SOTA
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Question AnsweringonTriviaQA
    F1· 2019-07-24
    83.6
    SOTA
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Question AnsweringonSearchQA
    F1· 2019-07-24
    84.8
    SOTA
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Coreference ResolutiononOntoNotes
    F1· 2019-07-24
    79.6
    best: 83.6 (Maverick_mes)
    SOTA
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Open-Domain Question AnsweringonSearchQA
    F1· 2019-07-24
    84.8
    SOTA
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Relation ExtractiononSemEval-2010 Task-8
    F1· 2022-08-20
    88.8
    best: 91.9 (SP)
    SPOT: Knowledge-Enhanced Language Representations for Information ExtractionarXiv:2208.09625
  • Relation ExtractiononTACRED
    F1· 2019-07-24
    70.8
    best: 86.6 (RAG4RE)
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Relation ClassificationonTACRED
    F1· 2019-07-24
    70.8
    best: 76.8 (DeepStruct multi-task w/ finetune)
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Question AnsweringonNewsQA
    F1· 2019-07-24
    73.6
    best: 94.01 (Riple/Saanvi-v0.5-DeepAnalysis)
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Question AnsweringonSQuAD2.0 dev
    F1· 2019-07-24
    86.8
    best: 90.6 (XLNet (single model))
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Question AnsweringonSQuAD2.0
    EM· 2019-07-24
    85.7
    best: 90.939 (IE-Net (ensemble))
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Question AnsweringonSQuAD2.0
    F1· 2019-07-24
    88.7
    best: 93.214 (IE-Net (ensemble))
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Natural Language InferenceonMultiNLI
    Matched· 2019-07-24
    88.1
    best: 92.6 (Turing NLR v5 XXL 5.4B (fine-tuned))
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Semantic Textual SimilarityonSTS Benchmark
    Pearson Correlation· 2019-07-24
    0.899
    best: 0.929 (MT-DNN-SMART)
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Semantic Textual SimilarityonQuora Question Pairs
    Accuracy· 2019-07-24
    89.5
    best: 92.4 (data2vec)
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Semantic Textual SimilarityonQuora Question Pairs
    F1· 2019-07-24
    71.9
    best: 90.7 (ALICE)
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Sentiment AnalysisonSST-2 Binary classification
    Accuracy· 2019-07-24
    94.8
    best: 97.5 (T5-11B)
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Paraphrase IdentificationonQuora Question Pairs
    Accuracy· 2019-07-24
    89.5
    best: 92.4 (data2vec)
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529
  • Paraphrase IdentificationonQuora Question Pairs
    F1· 2019-07-24
    71.9
    best: 90.7 (ALICE)
    SpanBERT: Improving Pre-training by Representing and Predicting SpansarXiv:1907.10529