{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:XMTRJKNEQTST7TUGWOJC2K7HBX","short_pith_number":"pith:XMTRJKNE","schema_version":"1.0","canonical_sha256":"bb2714a9a484e53fce86b3922d2be70de9dbb29709a5f8e4b0bbf419f05f643d","source":{"kind":"arxiv","id":"1909.13788","version":3},"attestation_state":"computed","paper":{"title":"Revisiting Self-Training for Neural Sequence Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jiajun Shen, Jiatao Gu, Junxian He, Marc'Aurelio Ranzato","submitted_at":"2019-09-30T15:30:00Z","abstract_excerpt":"Self-training is one of the earliest and simplest semi-supervised methods. The key idea is to augment the original labeled dataset with unlabeled data paired with the model's prediction (i.e. the pseudo-parallel data). While self-training has been extensively studied on classification problems, in complex sequence generation tasks (e.g. machine translation) it is still unclear how self-training works due to the compositionality of the target space. In this work, we first empirically show that self-training is able to decently improve the supervised baseline on neural sequence generation tasks."},"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":"1909.13788","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-09-30T15:30:00Z","cross_cats_sorted":["cs.CL","stat.ML"],"title_canon_sha256":"86916147f13cef644c79adc062884796e661627c9c2bfac273905487f8167d8b","abstract_canon_sha256":"8568a511fdd26496e6a2e0f1ffc0dae24c49bec520836456f96cda74eb596423"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:43:55.050065Z","signature_b64":"RffeAiqEgM6S0jm52aWyt+7ZaYmvbl0wch/Z1xrU8cM96rLaDCajE2UY1AkCr4NSoyUXlB/9URfoSbErgWk9BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bb2714a9a484e53fce86b3922d2be70de9dbb29709a5f8e4b0bbf419f05f643d","last_reissued_at":"2026-07-05T01:43:55.049584Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:43:55.049584Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Revisiting Self-Training for Neural Sequence Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jiajun Shen, Jiatao Gu, Junxian He, Marc'Aurelio Ranzato","submitted_at":"2019-09-30T15:30:00Z","abstract_excerpt":"Self-training is one of the earliest and simplest semi-supervised methods. The key idea is to augment the original labeled dataset with unlabeled data paired with the model's prediction (i.e. the pseudo-parallel data). While self-training has been extensively studied on classification problems, in complex sequence generation tasks (e.g. machine translation) it is still unclear how self-training works due to the compositionality of the target space. In this work, we first empirically show that self-training is able to decently improve the supervised baseline on neural sequence generation tasks."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1909.13788","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/1909.13788/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":"1909.13788","created_at":"2026-07-05T01:43:55.049648+00:00"},{"alias_kind":"arxiv_version","alias_value":"1909.13788v3","created_at":"2026-07-05T01:43:55.049648+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1909.13788","created_at":"2026-07-05T01:43:55.049648+00:00"},{"alias_kind":"pith_short_12","alias_value":"XMTRJKNEQTST","created_at":"2026-07-05T01:43:55.049648+00:00"},{"alias_kind":"pith_short_16","alias_value":"XMTRJKNEQTST7TUG","created_at":"2026-07-05T01:43:55.049648+00:00"},{"alias_kind":"pith_short_8","alias_value":"XMTRJKNE","created_at":"2026-07-05T01:43:55.049648+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2308.08998","citing_title":"Reinforced Self-Training (ReST) for Language Modeling","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XMTRJKNEQTST7TUGWOJC2K7HBX","json":"https://pith.science/pith/XMTRJKNEQTST7TUGWOJC2K7HBX.json","graph_json":"https://pith.science/api/pith-number/XMTRJKNEQTST7TUGWOJC2K7HBX/graph.json","events_json":"https://pith.science/api/pith-number/XMTRJKNEQTST7TUGWOJC2K7HBX/events.json","paper":"https://pith.science/paper/XMTRJKNE"},"agent_actions":{"view_html":"https://pith.science/pith/XMTRJKNEQTST7TUGWOJC2K7HBX","download_json":"https://pith.science/pith/XMTRJKNEQTST7TUGWOJC2K7HBX.json","view_paper":"https://pith.science/paper/XMTRJKNE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1909.13788&json=true","fetch_graph":"https://pith.science/api/pith-number/XMTRJKNEQTST7TUGWOJC2K7HBX/graph.json","fetch_events":"https://pith.science/api/pith-number/XMTRJKNEQTST7TUGWOJC2K7HBX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XMTRJKNEQTST7TUGWOJC2K7HBX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XMTRJKNEQTST7TUGWOJC2K7HBX/action/storage_attestation","attest_author":"https://pith.science/pith/XMTRJKNEQTST7TUGWOJC2K7HBX/action/author_attestation","sign_citation":"https://pith.science/pith/XMTRJKNEQTST7TUGWOJC2K7HBX/action/citation_signature","submit_replication":"https://pith.science/pith/XMTRJKNEQTST7TUGWOJC2K7HBX/action/replication_record"}},"created_at":"2026-07-05T01:43:55.049648+00:00","updated_at":"2026-07-05T01:43:55.049648+00:00"}