{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:F5NIT36WJDYOKSWEY7RKHUKQ7A","short_pith_number":"pith:F5NIT36W","schema_version":"1.0","canonical_sha256":"2f5a89efd648f0e54ac4c7e2a3d150f8287d2c48d551327bce4103bba161351f","source":{"kind":"arxiv","id":"2003.02245","version":2},"attestation_state":"computed","paper":{"title":"Data Augmentation using Pre-trained Transformer Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Ashutosh Choudhary, Eunah Cho, Varun Kumar","submitted_at":"2020-03-04T18:35:19Z","abstract_excerpt":"Language model based pre-trained models such as BERT have provided significant gains across different NLP tasks. In this paper, we study different types of transformer based pre-trained models such as auto-regressive models (GPT-2), auto-encoder models (BERT), and seq2seq models (BART) for conditional data augmentation. We show that prepending the class labels to text sequences provides a simple yet effective way to condition the pre-trained models for data augmentation. Additionally, on three classification benchmarks, pre-trained Seq2Seq model outperforms other data augmentation methods in a"},"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":"2003.02245","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-03-04T18:35:19Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"582130dc7608dbfc6dde4539057de32aa3090b269e0d974bcd7812a5c6e54eba","abstract_canon_sha256":"c793304c7865876f107f1e0ead79bd43e6e773b42af622ad85c4420255be987e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:10:59.584339Z","signature_b64":"PhSpr1VtZEG+r8tizmwKHqQp8rdO1MSKIhymFt9S2J0S1UZCBC+wUct0bhH6SdAiREjtotuCh2hx+grrMF4bAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2f5a89efd648f0e54ac4c7e2a3d150f8287d2c48d551327bce4103bba161351f","last_reissued_at":"2026-07-05T02:10:59.583853Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:10:59.583853Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Data Augmentation using Pre-trained Transformer Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Ashutosh Choudhary, Eunah Cho, Varun Kumar","submitted_at":"2020-03-04T18:35:19Z","abstract_excerpt":"Language model based pre-trained models such as BERT have provided significant gains across different NLP tasks. In this paper, we study different types of transformer based pre-trained models such as auto-regressive models (GPT-2), auto-encoder models (BERT), and seq2seq models (BART) for conditional data augmentation. We show that prepending the class labels to text sequences provides a simple yet effective way to condition the pre-trained models for data augmentation. Additionally, on three classification benchmarks, pre-trained Seq2Seq model outperforms other data augmentation methods in a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2003.02245","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/2003.02245/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":"2003.02245","created_at":"2026-07-05T02:10:59.583913+00:00"},{"alias_kind":"arxiv_version","alias_value":"2003.02245v2","created_at":"2026-07-05T02:10:59.583913+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2003.02245","created_at":"2026-07-05T02:10:59.583913+00:00"},{"alias_kind":"pith_short_12","alias_value":"F5NIT36WJDYO","created_at":"2026-07-05T02:10:59.583913+00:00"},{"alias_kind":"pith_short_16","alias_value":"F5NIT36WJDYOKSWE","created_at":"2026-07-05T02:10:59.583913+00:00"},{"alias_kind":"pith_short_8","alias_value":"F5NIT36W","created_at":"2026-07-05T02:10:59.583913+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20369","citing_title":"CATCH-ME if you RAG: a dataset of Contextually Annotated multi-Turn Counterspeech against Hate and Misinformation Exchanges","ref_index":290,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22435","citing_title":"Assisted Counterspeech Writing at the Crossroads of Hate Speech and Misinformation","ref_index":280,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/F5NIT36WJDYOKSWEY7RKHUKQ7A","json":"https://pith.science/pith/F5NIT36WJDYOKSWEY7RKHUKQ7A.json","graph_json":"https://pith.science/api/pith-number/F5NIT36WJDYOKSWEY7RKHUKQ7A/graph.json","events_json":"https://pith.science/api/pith-number/F5NIT36WJDYOKSWEY7RKHUKQ7A/events.json","paper":"https://pith.science/paper/F5NIT36W"},"agent_actions":{"view_html":"https://pith.science/pith/F5NIT36WJDYOKSWEY7RKHUKQ7A","download_json":"https://pith.science/pith/F5NIT36WJDYOKSWEY7RKHUKQ7A.json","view_paper":"https://pith.science/paper/F5NIT36W","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2003.02245&json=true","fetch_graph":"https://pith.science/api/pith-number/F5NIT36WJDYOKSWEY7RKHUKQ7A/graph.json","fetch_events":"https://pith.science/api/pith-number/F5NIT36WJDYOKSWEY7RKHUKQ7A/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/F5NIT36WJDYOKSWEY7RKHUKQ7A/action/timestamp_anchor","attest_storage":"https://pith.science/pith/F5NIT36WJDYOKSWEY7RKHUKQ7A/action/storage_attestation","attest_author":"https://pith.science/pith/F5NIT36WJDYOKSWEY7RKHUKQ7A/action/author_attestation","sign_citation":"https://pith.science/pith/F5NIT36WJDYOKSWEY7RKHUKQ7A/action/citation_signature","submit_replication":"https://pith.science/pith/F5NIT36WJDYOKSWEY7RKHUKQ7A/action/replication_record"}},"created_at":"2026-07-05T02:10:59.583913+00:00","updated_at":"2026-07-05T02:10:59.583913+00:00"}