{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:XIAQVYAFMVGB442CUL5PMGROBO","short_pith_number":"pith:XIAQVYAF","schema_version":"1.0","canonical_sha256":"ba010ae005654c1e7342a2faf61a2e0bb26884949397e361b4d37589ad1e54e6","source":{"kind":"arxiv","id":"2005.12766","version":2},"attestation_state":"computed","paper":{"title":"CERT: Contrastive Self-supervised Learning for Language Understanding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CL","authors_text":"Hongchao Fang, Jiayuan Ding, Meng Zhou, Pengtao Xie, Sicheng Wang","submitted_at":"2020-05-16T16:20:38Z","abstract_excerpt":"Pretrained language models such as BERT, GPT have shown great effectiveness in language understanding. The auxiliary predictive tasks in existing pretraining approaches are mostly defined on tokens, thus may not be able to capture sentence-level semantics very well. To address this issue, we propose CERT: Contrastive self-supervised Encoder Representations from Transformers, which pretrains language representation models using contrastive self-supervised learning at the sentence level. CERT creates augmentations of original sentences using back-translation. Then it finetunes a pretrained langu"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2005.12766","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-05-16T16:20:38Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"dfd64f171c546c52d796aec7c46a3adcfa311dc5cbcb79577606e593be86754c","abstract_canon_sha256":"66cf76ffd12c1dfa9a28b71cec6b394e59b5db0898667bf3df4cd967ef5a487b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:11:14.036439Z","signature_b64":"cxt16FQEP6MRojddMKR9N1KTr2O6uBbNZr16W+Wa6YprBz6B049huJsNlsSMlyQg0bXYGnTJQ6EYbaWKegifAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ba010ae005654c1e7342a2faf61a2e0bb26884949397e361b4d37589ad1e54e6","last_reissued_at":"2026-07-05T01:11:14.036034Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:11:14.036034Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CERT: Contrastive Self-supervised Learning for Language Understanding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CL","authors_text":"Hongchao Fang, Jiayuan Ding, Meng Zhou, Pengtao Xie, Sicheng Wang","submitted_at":"2020-05-16T16:20:38Z","abstract_excerpt":"Pretrained language models such as BERT, GPT have shown great effectiveness in language understanding. The auxiliary predictive tasks in existing pretraining approaches are mostly defined on tokens, thus may not be able to capture sentence-level semantics very well. To address this issue, we propose CERT: Contrastive self-supervised Encoder Representations from Transformers, which pretrains language representation models using contrastive self-supervised learning at the sentence level. CERT creates augmentations of original sentences using back-translation. Then it finetunes a pretrained langu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2005.12766","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2005.12766/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2005.12766","created_at":"2026-07-05T01:11:14.036092+00:00"},{"alias_kind":"arxiv_version","alias_value":"2005.12766v2","created_at":"2026-07-05T01:11:14.036092+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2005.12766","created_at":"2026-07-05T01:11:14.036092+00:00"},{"alias_kind":"pith_short_12","alias_value":"XIAQVYAFMVGB","created_at":"2026-07-05T01:11:14.036092+00:00"},{"alias_kind":"pith_short_16","alias_value":"XIAQVYAFMVGB442C","created_at":"2026-07-05T01:11:14.036092+00:00"},{"alias_kind":"pith_short_8","alias_value":"XIAQVYAF","created_at":"2026-07-05T01:11:14.036092+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.04061","citing_title":"Intra-Modal Neighbors Never Lie: Rectifying Inter-Modal Noisy Correspondence via Graph-Based Intra-Modal Reasoning","ref_index":91,"is_internal_anchor":false},{"citing_arxiv_id":"2606.27747","citing_title":"UNICS: Multilingual Code Search via Unified Pseudocode and Contrastive Transfer Learning","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2208.03299","citing_title":"Atlas: Few-shot Learning with Retrieval Augmented Language Models","ref_index":112,"is_internal_anchor":false},{"citing_arxiv_id":"2201.10005","citing_title":"Text and Code Embeddings by Contrastive Pre-Training","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2507.13334","citing_title":"A Survey of Context Engineering for Large Language Models","ref_index":267,"is_internal_anchor":false},{"citing_arxiv_id":"2112.09118","citing_title":"Unsupervised Dense Information Retrieval with Contrastive Learning","ref_index":128,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XIAQVYAFMVGB442CUL5PMGROBO","json":"https://pith.science/pith/XIAQVYAFMVGB442CUL5PMGROBO.json","graph_json":"https://pith.science/api/pith-number/XIAQVYAFMVGB442CUL5PMGROBO/graph.json","events_json":"https://pith.science/api/pith-number/XIAQVYAFMVGB442CUL5PMGROBO/events.json","paper":"https://pith.science/paper/XIAQVYAF"},"agent_actions":{"view_html":"https://pith.science/pith/XIAQVYAFMVGB442CUL5PMGROBO","download_json":"https://pith.science/pith/XIAQVYAFMVGB442CUL5PMGROBO.json","view_paper":"https://pith.science/paper/XIAQVYAF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2005.12766&json=true","fetch_graph":"https://pith.science/api/pith-number/XIAQVYAFMVGB442CUL5PMGROBO/graph.json","fetch_events":"https://pith.science/api/pith-number/XIAQVYAFMVGB442CUL5PMGROBO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XIAQVYAFMVGB442CUL5PMGROBO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XIAQVYAFMVGB442CUL5PMGROBO/action/storage_attestation","attest_author":"https://pith.science/pith/XIAQVYAFMVGB442CUL5PMGROBO/action/author_attestation","sign_citation":"https://pith.science/pith/XIAQVYAFMVGB442CUL5PMGROBO/action/citation_signature","submit_replication":"https://pith.science/pith/XIAQVYAFMVGB442CUL5PMGROBO/action/replication_record"}},"created_at":"2026-07-05T01:11:14.036092+00:00","updated_at":"2026-07-05T01:11:14.036092+00:00"}