CLEAR: Contrastive Learning for Sentence Representation

Benchmark Model Rank Results
linguistic-acceptability-on-colaMLM+ del-span+ reorderAccuracy: 64.3%
natural-language-inference-on-qnliMLM+ subs+ del-spanAccuracy: 93.4%
natural-language-inference-on-rteMLM+ del-spanAccuracy: 79.8%
question-answering-on-quora-question-pairsMLM+ subs+ del-spanAccuracy: 90.3%
semantic-textual-similarity-on-mrpcMLM+ del-word+ reorderAccuracy: 90.6%
semantic-textual-similarity-on-sts-benchmarkMLM+ del-wordPearson Correlation: 0.905
sentiment-analysis-on-sst-2-binaryMLM+ del-word+ reorderAccuracy: 94.5