{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:MMMN62J6GXRARUV6N53JR5YIVV","short_pith_number":"pith:MMMN62J6","schema_version":"1.0","canonical_sha256":"6318df693e35e208d2be6f7698f708ad7bebdcae1944f389ad2a7e2d8eff69ba","source":{"kind":"arxiv","id":"2512.11388","version":1},"attestation_state":"computed","paper":{"title":"Improving Translation Quality by Selecting Better Data for LLM Fine-Tuning: A Comparative Analysis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Felipe Ribeiro Fujita de Mello, Hideyuki Takada","submitted_at":"2025-12-12T08:59:27Z","abstract_excerpt":"We investigated the impact of data selection on machine translation fine-tuning for open LLMs. Using Japanese-English corpora, we compare five selectors: TF-IDF, COMET Kiwi, QuRate, FD-Score, and random selection, under controlled training conditions. We observed that semantic selectors consistently outperform lexical and geometry-based heuristics, and that even when the selected data differ by less than 3%, the impact on model performance is substantial, underscoring the sensitivity of fine-tuning to data quality."},"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":"2512.11388","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-12-12T08:59:27Z","cross_cats_sorted":[],"title_canon_sha256":"6c606425dd6c7c9a91ec506fb8d5196ba05591c9fb9ea0f30f75c0e1c0439da3","abstract_canon_sha256":"6bcaaf7ba0f5c7661cb87928e1d305033cda6acc8c4a1c80c527a6ed1536780b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-07T02:19:43.620247Z","signature_b64":"/GsC2exDJuRmoQqPsCgV0ch3hDA524g8N8LOdmG2HqgF5KN+4lJOqiNJiYdClblF7IAzNOURa+IoELxEUJDsBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6318df693e35e208d2be6f7698f708ad7bebdcae1944f389ad2a7e2d8eff69ba","last_reissued_at":"2026-07-07T02:19:43.619362Z","signature_status":"signed_v1","first_computed_at":"2026-07-07T02:19:43.619362Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Improving Translation Quality by Selecting Better Data for LLM Fine-Tuning: A Comparative Analysis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Felipe Ribeiro Fujita de Mello, Hideyuki Takada","submitted_at":"2025-12-12T08:59:27Z","abstract_excerpt":"We investigated the impact of data selection on machine translation fine-tuning for open LLMs. Using Japanese-English corpora, we compare five selectors: TF-IDF, COMET Kiwi, QuRate, FD-Score, and random selection, under controlled training conditions. We observed that semantic selectors consistently outperform lexical and geometry-based heuristics, and that even when the selected data differ by less than 3%, the impact on model performance is substantial, underscoring the sensitivity of fine-tuning to data quality."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.11388","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/2512.11388/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":"2512.11388","created_at":"2026-07-07T02:19:43.619472+00:00"},{"alias_kind":"arxiv_version","alias_value":"2512.11388v1","created_at":"2026-07-07T02:19:43.619472+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.11388","created_at":"2026-07-07T02:19:43.619472+00:00"},{"alias_kind":"pith_short_12","alias_value":"MMMN62J6GXRA","created_at":"2026-07-07T02:19:43.619472+00:00"},{"alias_kind":"pith_short_16","alias_value":"MMMN62J6GXRARUV6","created_at":"2026-07-07T02:19:43.619472+00:00"},{"alias_kind":"pith_short_8","alias_value":"MMMN62J6","created_at":"2026-07-07T02:19:43.619472+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/MMMN62J6GXRARUV6N53JR5YIVV","json":"https://pith.science/pith/MMMN62J6GXRARUV6N53JR5YIVV.json","graph_json":"https://pith.science/api/pith-number/MMMN62J6GXRARUV6N53JR5YIVV/graph.json","events_json":"https://pith.science/api/pith-number/MMMN62J6GXRARUV6N53JR5YIVV/events.json","paper":"https://pith.science/paper/MMMN62J6"},"agent_actions":{"view_html":"https://pith.science/pith/MMMN62J6GXRARUV6N53JR5YIVV","download_json":"https://pith.science/pith/MMMN62J6GXRARUV6N53JR5YIVV.json","view_paper":"https://pith.science/paper/MMMN62J6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2512.11388&json=true","fetch_graph":"https://pith.science/api/pith-number/MMMN62J6GXRARUV6N53JR5YIVV/graph.json","fetch_events":"https://pith.science/api/pith-number/MMMN62J6GXRARUV6N53JR5YIVV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MMMN62J6GXRARUV6N53JR5YIVV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MMMN62J6GXRARUV6N53JR5YIVV/action/storage_attestation","attest_author":"https://pith.science/pith/MMMN62J6GXRARUV6N53JR5YIVV/action/author_attestation","sign_citation":"https://pith.science/pith/MMMN62J6GXRARUV6N53JR5YIVV/action/citation_signature","submit_replication":"https://pith.science/pith/MMMN62J6GXRARUV6N53JR5YIVV/action/replication_record"}},"created_at":"2026-07-07T02:19:43.619472+00:00","updated_at":"2026-07-07T02:19:43.619472+00:00"}