A Broad-Coverage Challenge Corpus for Sentence Understanding through Inference
About
This paper introduces the Multi-Genre Natural Language Inference (MultiNLI) corpus, a dataset designed for use in the development and evaluation of machine learning models for sentence understanding. In addition to being one of the largest corpora available for the task of NLI, at 433k examples, this corpus improves upon available resources in its coverage: it offers data from ten distinct genres of written and spoken English--making it possible to evaluate systems on nearly the full complexity of the language--and it offers an explicit setting for the evaluation of cross-genre domain adaptation.
Adina Williams, Nikita Nangia, Samuel R. Bowman• 2017
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Natural Language Inference | SNLI (test) | Accuracy81.5 | 681 | |
| Sentiment Classification | Stanford Sentiment Treebank SST-2 (test) | Accuracy87.2 | 99 | |
| Natural Language Inference | MultiNLI matched (test) | Accuracy67.5 | 65 | |
| Natural Language Inference | MultiNLI Mismatched | Accuracy69.4 | 60 | |
| Natural Language Inference | MultiNLI mismatched (test) | Accuracy67.6 | 56 | |
| Factual Consistency Evaluation | SummaC | CGS46 | 52 | |
| Natural Language Inference | MultiNLI Matched | Accuracy69.8 | 49 | |
| Factual Consistency Evaluation | QAGS XSUM | Spearman Correlation0.7 | 39 | |
| Factual Consistency Evaluation | QAGS CNNDM | Spearman Correlation-16.4 | 38 | |
| Factual Consistency Evaluation | TRUE benchmark | PAWS (AUC-ROC)83.11 | 37 |
Showing 10 of 48 rows