{"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/diffcse-difference-based-contrastive-learning","title":"DiffCSE: Difference-based Contrastive Learning for Sentence Embeddings","arxiv_id":"2204.10298","date":"2022-04-21","proceeding":"NAACL 2022 7","authors":["Yung-Sung Chuang","Rumen Dangovski","Hongyin Luo","Yang Zhang","Shiyu Chang","Marin Soljačić","Shang-Wen Li","Wen-tau Yih","Yoon Kim","James Glass"],"abstract":"We propose DiffCSE, an unsupervised contrastive learning framework for learning sentence embeddings. DiffCSE learns sentence embeddings that are sensitive to the difference between the original sentence and an edited sentence, where the edited sentence is obtained by stochastically masking out the original sentence and then sampling from a masked language model. We show that DiffSCE is an instance of equivariant contrastive learning (Dangovski et al., 2021), which generalizes contrastive learning and learns representations that are insensitive to certain types of augmentations and sensitive to other \"harmful\" types of augmentations. Our experiments show that DiffCSE achieves state-of-the-art results among unsupervised sentence representation learning methods, outperforming unsupervised SimCSE by 2.3 absolute points on semantic textual similarity tasks.","url_abs":"https://arxiv.org/abs/2204.10298v1","url_pdf":"https://arxiv.org/pdf/2204.10298v1.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":[{"paper_slug":"diffcse-difference-based-contrastive-learning","repo_url":"https://github.com/voidism/diffcse","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"jax","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"contrastive-learning","task_name":"Contrastive Learning"},{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"representation-learning","task_name":"Representation Learning"},{"task_slug":"semantic-textual-similarity","task_name":"Semantic Textual Similarity"},{"task_slug":"sentence","task_name":"Sentence"},{"task_slug":"sentence-embeddings","task_name":"Sentence Embeddings"}],"methods":[{"method_slug":"contrastive-learning","method_name":"Contrastive Learning"},{"method_slug":"simcse","method_name":"SimCSE"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/semantic-textual-similarity-on-sts12","task":"Semantic Textual Similarity","dataset":"STS12","model":"DiffCSE-BERT-base","rank_in_archive_order":13,"of":20,"metrics":{"Spearman Correlation":"0.7228"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-textual-similarity-on-sts12","task":"Semantic Textual Similarity","dataset":"STS12","model":"DiffCSE-RoBERTa-base","rank_in_archive_order":16,"of":20,"metrics":{"Spearman Correlation":"0.7005"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-textual-similarity-on-sts13","task":"Semantic Textual Similarity","dataset":"STS13","model":"DiffCSE-BERT-base","rank_in_archive_order":13,"of":22,"metrics":{"Spearman Correlation":"0.8443"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-textual-similarity-on-sts13","task":"Semantic Textual Similarity","dataset":"STS13","model":"DiffCSE-RoBERTa-base","rank_in_archive_order":14,"of":22,"metrics":{"Spearman Correlation":"0.8343"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-textual-similarity-on-sts14","task":"Semantic Textual Similarity","dataset":"STS14","model":"DiffCSE-BERT-base","rank_in_archive_order":14,"of":21,"metrics":{"Spearman Correlation":"0.7647"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-textual-similarity-on-sts14","task":"Semantic Textual Similarity","dataset":"STS14","model":"DiffCSE-RoBERTa-base","rank_in_archive_order":15,"of":21,"metrics":{"Spearman Correlation":"0.7549"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-textual-similarity-on-sts15","task":"Semantic Textual Similarity","dataset":"STS15","model":"DiffCSE-BERT-base","rank_in_archive_order":13,"of":20,"metrics":{"Spearman Correlation":"0.8390"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-textual-similarity-on-sts15","task":"Semantic Textual Similarity","dataset":"STS15","model":"DiffCSE-RoBERTa-base","rank_in_archive_order":14,"of":20,"metrics":{"Spearman Correlation":"0.8281"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-textual-similarity-on-sts16","task":"Semantic Textual Similarity","dataset":"STS16","model":"DiffCSE-RoBERTa-base","rank_in_archive_order":13,"of":20,"metrics":{"Spearman Correlation":"0.8212"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-textual-similarity-on-sts16","task":"Semantic Textual Similarity","dataset":"STS16","model":"DiffCSE-BERT-base","rank_in_archive_order":14,"of":20,"metrics":{"Spearman Correlation":"0.8054"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2204.10298","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2204.10298"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/voidism/diffcse","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_draft_wrong":2,"unverified":2},"by_repo_kind":{"official":{"samples":4,"ran":2,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"d4903874fdec3763","entry":"create_position_ids_from_input_ids","repo":"voidism/diffcse","repo_kind":"official","path":"modeling_roberta.py","file_url":"https://github.com/voidism/diffcse/blob/HEAD/modeling_roberta.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"d4903874fdec3763"}},{"code_sha256_prefix":"cb67cd5ac977480d","entry":"sentemb_forward","repo":"voidism/diffcse","repo_kind":"official","path":"diffcse/models.py","file_url":"https://github.com/voidism/diffcse/blob/HEAD/diffcse/models.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"cb67cd5ac977480d"}},{"code_sha256_prefix":"c4cccc3c604c8c75","entry":"cl_forward","repo":"voidism/diffcse","repo_kind":"official","path":"diffcse/models.py","file_url":"https://github.com/voidism/diffcse/blob/HEAD/diffcse/models.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c4cccc3c604c8c75"}},{"code_sha256_prefix":"26be70dca3249c0b","entry":"load_tf_weights_in_bert","repo":"voidism/diffcse","repo_kind":"official","path":"modeling_bert.py","file_url":"https://github.com/voidism/diffcse/blob/HEAD/modeling_bert.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"26be70dca3249c0b"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}