{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:RWVMBIX5IGJQ5PHIZCXY35PJVK","short_pith_number":"pith:RWVMBIX5","schema_version":"1.0","canonical_sha256":"8daac0a2fd41930ebce8c8af8df5e9aaa9a9753eccfd4c6a486e1d969ba7bae1","source":{"kind":"arxiv","id":"2104.06979","version":3},"attestation_state":"computed","paper":{"title":"TSDAE: Using Transformer-based Sequential Denoising Auto-Encoder for Unsupervised Sentence Embedding Learning","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Iryna Gurevych, Kexin Wang, Nils Reimers","submitted_at":"2021-04-14T17:02:18Z","abstract_excerpt":"Learning sentence embeddings often requires a large amount of labeled data. However, for most tasks and domains, labeled data is seldom available and creating it is expensive. In this work, we present a new state-of-the-art unsupervised method based on pre-trained Transformers and Sequential Denoising Auto-Encoder (TSDAE) which outperforms previous approaches by up to 6.4 points. It can achieve up to 93.1% of the performance of in-domain supervised approaches. Further, we show that TSDAE is a strong domain adaptation and pre-training method for sentence embeddings, significantly outperforming "},"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":"2104.06979","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2021-04-14T17:02:18Z","cross_cats_sorted":[],"title_canon_sha256":"7cc752b4ac847069400a9539d9585b609463f2e096a315d9b43cd2320eab8dcf","abstract_canon_sha256":"39e0d2fc3a2b79b30f6443ce88eee4c88045e30393af282e0b232b34784c7e21"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:13:02.904098Z","signature_b64":"2HpA+EZGl2mbP7vg20THdXH6XWI0dn25/Sm8kbAqyusluEsM/ChjfZaBoOR1tx9+b1B3pR7XodO4fa40j1tlBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8daac0a2fd41930ebce8c8af8df5e9aaa9a9753eccfd4c6a486e1d969ba7bae1","last_reissued_at":"2026-07-05T03:13:02.903661Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:13:02.903661Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TSDAE: Using Transformer-based Sequential Denoising Auto-Encoder for Unsupervised Sentence Embedding Learning","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Iryna Gurevych, Kexin Wang, Nils Reimers","submitted_at":"2021-04-14T17:02:18Z","abstract_excerpt":"Learning sentence embeddings often requires a large amount of labeled data. However, for most tasks and domains, labeled data is seldom available and creating it is expensive. In this work, we present a new state-of-the-art unsupervised method based on pre-trained Transformers and Sequential Denoising Auto-Encoder (TSDAE) which outperforms previous approaches by up to 6.4 points. It can achieve up to 93.1% of the performance of in-domain supervised approaches. Further, we show that TSDAE is a strong domain adaptation and pre-training method for sentence embeddings, significantly outperforming "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.06979","kind":"arxiv","version":3},"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/2104.06979/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":"2104.06979","created_at":"2026-07-05T03:13:02.903719+00:00"},{"alias_kind":"arxiv_version","alias_value":"2104.06979v3","created_at":"2026-07-05T03:13:02.903719+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.06979","created_at":"2026-07-05T03:13:02.903719+00:00"},{"alias_kind":"pith_short_12","alias_value":"RWVMBIX5IGJQ","created_at":"2026-07-05T03:13:02.903719+00:00"},{"alias_kind":"pith_short_16","alias_value":"RWVMBIX5IGJQ5PHI","created_at":"2026-07-05T03:13:02.903719+00:00"},{"alias_kind":"pith_short_8","alias_value":"RWVMBIX5","created_at":"2026-07-05T03:13:02.903719+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.08184","citing_title":"TextEconomizer: Enhancing Lossy Text Compression with Denoising Transformers and Entropy Coding","ref_index":63,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RWVMBIX5IGJQ5PHIZCXY35PJVK","json":"https://pith.science/pith/RWVMBIX5IGJQ5PHIZCXY35PJVK.json","graph_json":"https://pith.science/api/pith-number/RWVMBIX5IGJQ5PHIZCXY35PJVK/graph.json","events_json":"https://pith.science/api/pith-number/RWVMBIX5IGJQ5PHIZCXY35PJVK/events.json","paper":"https://pith.science/paper/RWVMBIX5"},"agent_actions":{"view_html":"https://pith.science/pith/RWVMBIX5IGJQ5PHIZCXY35PJVK","download_json":"https://pith.science/pith/RWVMBIX5IGJQ5PHIZCXY35PJVK.json","view_paper":"https://pith.science/paper/RWVMBIX5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2104.06979&json=true","fetch_graph":"https://pith.science/api/pith-number/RWVMBIX5IGJQ5PHIZCXY35PJVK/graph.json","fetch_events":"https://pith.science/api/pith-number/RWVMBIX5IGJQ5PHIZCXY35PJVK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RWVMBIX5IGJQ5PHIZCXY35PJVK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RWVMBIX5IGJQ5PHIZCXY35PJVK/action/storage_attestation","attest_author":"https://pith.science/pith/RWVMBIX5IGJQ5PHIZCXY35PJVK/action/author_attestation","sign_citation":"https://pith.science/pith/RWVMBIX5IGJQ5PHIZCXY35PJVK/action/citation_signature","submit_replication":"https://pith.science/pith/RWVMBIX5IGJQ5PHIZCXY35PJVK/action/replication_record"}},"created_at":"2026-07-05T03:13:02.903719+00:00","updated_at":"2026-07-05T03:13:02.903719+00:00"}