{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:S5TKTKM7TX7IPRSUJGGBTKS2D2","short_pith_number":"pith:S5TKTKM7","schema_version":"1.0","canonical_sha256":"9766a9a99f9dfe87c654498c19aa5a1eb1dc61c07293ce4c5f1a5b9fec3a845d","source":{"kind":"arxiv","id":"1908.04577","version":3},"attestation_state":"computed","paper":{"title":"StructBERT: Incorporating Language Structures into Pre-training for Deep Language Understanding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bin Bi, Chen Wu, Jiangnan Xia, Liwei Peng, Luo Si, Ming Yan, Wei Wang, Zuyi Bao","submitted_at":"2019-08-13T11:12:58Z","abstract_excerpt":"Recently, the pre-trained language model, BERT (and its robustly optimized version RoBERTa), has attracted a lot of attention in natural language understanding (NLU), and achieved state-of-the-art accuracy in various NLU tasks, such as sentiment classification, natural language inference, semantic textual similarity and question answering. Inspired by the linearization exploration work of Elman [8], we extend BERT to a new model, StructBERT, by incorporating language structures into pre-training. Specifically, we pre-train StructBERT with two auxiliary tasks to make the most of the sequential "},"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":"1908.04577","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-08-13T11:12:58Z","cross_cats_sorted":[],"title_canon_sha256":"a859c24f91f1354713647ca1a277a0479fe4ef529bffd31a71a13d24eb6c4b4c","abstract_canon_sha256":"39a7357eae8870c03147ab7d0cc909cfb26de91444eb137f12b3efe568b12188"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:07:37.186092Z","signature_b64":"C9uSWW+K8sO1gy+iwjnJBxe7gL7+R5wY+c5hHzBxmD9ivOSLYFg4DIguw0b5G6e7MU2jLwFtAH5Zw2JQSTy3Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9766a9a99f9dfe87c654498c19aa5a1eb1dc61c07293ce4c5f1a5b9fec3a845d","last_reissued_at":"2026-07-05T00:07:37.185643Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:07:37.185643Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"StructBERT: Incorporating Language Structures into Pre-training for Deep Language Understanding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bin Bi, Chen Wu, Jiangnan Xia, Liwei Peng, Luo Si, Ming Yan, Wei Wang, Zuyi Bao","submitted_at":"2019-08-13T11:12:58Z","abstract_excerpt":"Recently, the pre-trained language model, BERT (and its robustly optimized version RoBERTa), has attracted a lot of attention in natural language understanding (NLU), and achieved state-of-the-art accuracy in various NLU tasks, such as sentiment classification, natural language inference, semantic textual similarity and question answering. Inspired by the linearization exploration work of Elman [8], we extend BERT to a new model, StructBERT, by incorporating language structures into pre-training. Specifically, we pre-train StructBERT with two auxiliary tasks to make the most of the sequential "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1908.04577","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/1908.04577/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":"1908.04577","created_at":"2026-07-05T00:07:37.185697+00:00"},{"alias_kind":"arxiv_version","alias_value":"1908.04577v3","created_at":"2026-07-05T00:07:37.185697+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1908.04577","created_at":"2026-07-05T00:07:37.185697+00:00"},{"alias_kind":"pith_short_12","alias_value":"S5TKTKM7TX7I","created_at":"2026-07-05T00:07:37.185697+00:00"},{"alias_kind":"pith_short_16","alias_value":"S5TKTKM7TX7IPRSU","created_at":"2026-07-05T00:07:37.185697+00:00"},{"alias_kind":"pith_short_8","alias_value":"S5TKTKM7","created_at":"2026-07-05T00:07:37.185697+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2510.11233","citing_title":"CNSocialDepress: A Chinese Social Media Dataset for Depression Risk Detection and Structured Analysis","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2006.03654","citing_title":"DeBERTa: Decoding-enhanced BERT with Disentangled Attention","ref_index":32,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/S5TKTKM7TX7IPRSUJGGBTKS2D2","json":"https://pith.science/pith/S5TKTKM7TX7IPRSUJGGBTKS2D2.json","graph_json":"https://pith.science/api/pith-number/S5TKTKM7TX7IPRSUJGGBTKS2D2/graph.json","events_json":"https://pith.science/api/pith-number/S5TKTKM7TX7IPRSUJGGBTKS2D2/events.json","paper":"https://pith.science/paper/S5TKTKM7"},"agent_actions":{"view_html":"https://pith.science/pith/S5TKTKM7TX7IPRSUJGGBTKS2D2","download_json":"https://pith.science/pith/S5TKTKM7TX7IPRSUJGGBTKS2D2.json","view_paper":"https://pith.science/paper/S5TKTKM7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1908.04577&json=true","fetch_graph":"https://pith.science/api/pith-number/S5TKTKM7TX7IPRSUJGGBTKS2D2/graph.json","fetch_events":"https://pith.science/api/pith-number/S5TKTKM7TX7IPRSUJGGBTKS2D2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/S5TKTKM7TX7IPRSUJGGBTKS2D2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/S5TKTKM7TX7IPRSUJGGBTKS2D2/action/storage_attestation","attest_author":"https://pith.science/pith/S5TKTKM7TX7IPRSUJGGBTKS2D2/action/author_attestation","sign_citation":"https://pith.science/pith/S5TKTKM7TX7IPRSUJGGBTKS2D2/action/citation_signature","submit_replication":"https://pith.science/pith/S5TKTKM7TX7IPRSUJGGBTKS2D2/action/replication_record"}},"created_at":"2026-07-05T00:07:37.185697+00:00","updated_at":"2026-07-05T00:07:37.185697+00:00"}