{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/clear-contrastive-learning-for-sentence","title":"CLEAR: Contrastive Learning for Sentence Representation","arxiv_id":"2012.15466","date":"2020-12-31","proceeding":null,"authors":["Zhuofeng Wu","Sinong Wang","Jiatao Gu","Madian Khabsa","Fei Sun","Hao Ma"],"abstract":"Pre-trained language models have proven their unique powers in capturing implicit language features. However, most pre-training approaches focus on the word-level training objective, while sentence-level objectives are rarely studied. In this paper, we propose Contrastive LEArning for sentence Representation (CLEAR), which employs multiple sentence-level augmentation strategies in order to learn a noise-invariant sentence representation. These augmentations include word and span deletion, reordering, and substitution. Furthermore, we investigate the key reasons that make contrastive learning effective through numerous experiments. We observe that different sentence augmentations during pre-training lead to different performance improvements on various downstream tasks. Our approach is shown to outperform multiple existing methods on both SentEval and GLUE benchmarks.","url_abs":"https://arxiv.org/abs/2012.15466v1","url_pdf":"https://arxiv.org/pdf/2012.15466v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[],"tasks":[{"task_slug":"contrastive-learning","task_name":"Contrastive Learning"},{"task_slug":"linguistic-acceptability","task_name":"Linguistic Acceptability"},{"task_slug":"natural-language-inference","task_name":"Natural Language Inference"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"semantic-textual-similarity","task_name":"Semantic Textual Similarity"},{"task_slug":"sentence","task_name":"Sentence"},{"task_slug":"sentiment-analysis","task_name":"Sentiment Analysis"}],"methods":[{"method_slug":"contrastive-learning","method_name":"Contrastive Learning"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/linguistic-acceptability-on-cola","task":"Linguistic Acceptability","dataset":"CoLA","model":"MLM+ del-span+ reorder","rank_in_archive_order":26,"of":43,"metrics":{"Accuracy":"64.3%"},"uses_additional_data":false},{"leaderboard":"/sota/natural-language-inference-on-qnli","task":"Natural Language Inference","dataset":"QNLI","model":"MLM+ subs+ del-span","rank_in_archive_order":21,"of":43,"metrics":{"Accuracy":"93.4%"},"uses_additional_data":false},{"leaderboard":"/sota/natural-language-inference-on-rte","task":"Natural Language Inference","dataset":"RTE","model":"MLM+ del-span","rank_in_archive_order":37,"of":90,"metrics":{"Accuracy":"79.8%"},"uses_additional_data":false},{"leaderboard":"/sota/question-answering-on-quora-question-pairs","task":"Question Answering","dataset":"Quora Question Pairs","model":"MLM+ subs+ del-span","rank_in_archive_order":5,"of":19,"metrics":{"Accuracy":"90.3%"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-textual-similarity-on-mrpc","task":"Semantic Textual Similarity","dataset":"MRPC","model":"MLM+ del-word+ reorder","rank_in_archive_order":11,"of":45,"metrics":{"Accuracy":"90.6%"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-textual-similarity-on-sts-benchmark","task":"Semantic Textual Similarity","dataset":"STS Benchmark","model":"MLM+ del-word","rank_in_archive_order":18,"of":66,"metrics":{"Pearson Correlation":"0.905"},"uses_additional_data":false},{"leaderboard":"/sota/sentiment-analysis-on-sst-2-binary","task":"Sentiment Analysis","dataset":"SST-2 Binary classification","model":"MLM+ del-word+ reorder","rank_in_archive_order":35,"of":87,"metrics":{"Accuracy":"94.5"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2012.15466","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}