{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:PTKB7E67ONR2BMFXBPTGGRLNRC","short_pith_number":"pith:PTKB7E67","schema_version":"1.0","canonical_sha256":"7cd41f93df7363a0b0b70be663456d88a9a39484bce6d8982e1af291d2153fbd","source":{"kind":"arxiv","id":"2503.22582","version":1},"attestation_state":"computed","paper":{"title":"Beyond Vanilla Fine-Tuning: Leveraging Multistage, Multilingual, and Domain-Specific Methods for Low-Resource Machine Translation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"En-Shiun Annie Lee, Sanath Jayasena, Sarubi Thillainathan, Songchen Yuan, Surangika Ranathunga","submitted_at":"2025-03-28T16:30:28Z","abstract_excerpt":"Fine-tuning multilingual sequence-to-sequence large language models (msLLMs) has shown promise in developing neural machine translation (NMT) systems for low-resource languages (LRLs). However, conventional single-stage fine-tuning methods struggle in extremely low-resource NMT settings, where training data is very limited. This paper contributes to artificial intelligence by proposing two approaches for adapting msLLMs in these challenging scenarios: (1) continual pre-training (CPT), where the msLLM is further trained with domain-specific monolingual data to compensate for the under-represent"},"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":"2503.22582","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-28T16:30:28Z","cross_cats_sorted":[],"title_canon_sha256":"c5822a007e35a9e0da503d746637497da99c74c9571003c93f187c0abcda0ab8","abstract_canon_sha256":"f22d425e3f86d288a9927c38aad054b103e610a98e0e2c8adecb489d2f4ef827"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:41:07.076702Z","signature_b64":"fW0oexYx+5Tw50JHpRZbOT8I2AvwBO8ef6j9lSg5nR91nycq1GQuT0K0Hw1n8I++D8gOTQIzLEF0A+kwksZXAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7cd41f93df7363a0b0b70be663456d88a9a39484bce6d8982e1af291d2153fbd","last_reissued_at":"2026-07-05T10:41:07.076130Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:41:07.076130Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Vanilla Fine-Tuning: Leveraging Multistage, Multilingual, and Domain-Specific Methods for Low-Resource Machine Translation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"En-Shiun Annie Lee, Sanath Jayasena, Sarubi Thillainathan, Songchen Yuan, Surangika Ranathunga","submitted_at":"2025-03-28T16:30:28Z","abstract_excerpt":"Fine-tuning multilingual sequence-to-sequence large language models (msLLMs) has shown promise in developing neural machine translation (NMT) systems for low-resource languages (LRLs). However, conventional single-stage fine-tuning methods struggle in extremely low-resource NMT settings, where training data is very limited. This paper contributes to artificial intelligence by proposing two approaches for adapting msLLMs in these challenging scenarios: (1) continual pre-training (CPT), where the msLLM is further trained with domain-specific monolingual data to compensate for the under-represent"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.22582","kind":"arxiv","version":1},"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/2503.22582/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":"2503.22582","created_at":"2026-07-05T10:41:07.076190+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.22582v1","created_at":"2026-07-05T10:41:07.076190+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.22582","created_at":"2026-07-05T10:41:07.076190+00:00"},{"alias_kind":"pith_short_12","alias_value":"PTKB7E67ONR2","created_at":"2026-07-05T10:41:07.076190+00:00"},{"alias_kind":"pith_short_16","alias_value":"PTKB7E67ONR2BMFX","created_at":"2026-07-05T10:41:07.076190+00:00"},{"alias_kind":"pith_short_8","alias_value":"PTKB7E67","created_at":"2026-07-05T10:41:07.076190+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PTKB7E67ONR2BMFXBPTGGRLNRC","json":"https://pith.science/pith/PTKB7E67ONR2BMFXBPTGGRLNRC.json","graph_json":"https://pith.science/api/pith-number/PTKB7E67ONR2BMFXBPTGGRLNRC/graph.json","events_json":"https://pith.science/api/pith-number/PTKB7E67ONR2BMFXBPTGGRLNRC/events.json","paper":"https://pith.science/paper/PTKB7E67"},"agent_actions":{"view_html":"https://pith.science/pith/PTKB7E67ONR2BMFXBPTGGRLNRC","download_json":"https://pith.science/pith/PTKB7E67ONR2BMFXBPTGGRLNRC.json","view_paper":"https://pith.science/paper/PTKB7E67","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.22582&json=true","fetch_graph":"https://pith.science/api/pith-number/PTKB7E67ONR2BMFXBPTGGRLNRC/graph.json","fetch_events":"https://pith.science/api/pith-number/PTKB7E67ONR2BMFXBPTGGRLNRC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PTKB7E67ONR2BMFXBPTGGRLNRC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PTKB7E67ONR2BMFXBPTGGRLNRC/action/storage_attestation","attest_author":"https://pith.science/pith/PTKB7E67ONR2BMFXBPTGGRLNRC/action/author_attestation","sign_citation":"https://pith.science/pith/PTKB7E67ONR2BMFXBPTGGRLNRC/action/citation_signature","submit_replication":"https://pith.science/pith/PTKB7E67ONR2BMFXBPTGGRLNRC/action/replication_record"}},"created_at":"2026-07-05T10:41:07.076190+00:00","updated_at":"2026-07-05T10:41:07.076190+00:00"}