{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XU4523D2XXPNAQN2BF5I2OPI4P","short_pith_number":"pith:XU4523D2","schema_version":"1.0","canonical_sha256":"bd39dd6c7abdded041ba097a8d39e8e3d2d9e8913054315c4206cbf75671debd","source":{"kind":"arxiv","id":"2505.00114","version":1},"attestation_state":"computed","paper":{"title":"Fine-Tuning LLMs for Low-Resource Dialect Translation: The Case of Lebanese","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ali Chehab, Silvana Yakhni","submitted_at":"2025-04-30T18:33:53Z","abstract_excerpt":"This paper examines the effectiveness of Large Language Models (LLMs) in translating the low-resource Lebanese dialect, focusing on the impact of culturally authentic data versus larger translated datasets. We compare three fine-tuning approaches: Basic, contrastive, and grammar-hint tuning, using open-source Aya23 models. Experiments reveal that models fine-tuned on a smaller but culturally aware Lebanese dataset (LW) consistently outperform those trained on larger, non-native data. The best results were achieved through contrastive fine-tuning paired with contrastive prompting, which indicat"},"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":"2505.00114","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/publicdomain/zero/1.0/","primary_cat":"cs.CL","submitted_at":"2025-04-30T18:33:53Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"63310288ffe3eb231b5852182fafcf5fefbfc8c4f3321feaa4de3157a5f4b041","abstract_canon_sha256":"6ba23d282bc7ffd87b87b303bc471aa0bc680362eb534408e39a87dcf0c51b47"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:56:41.937859Z","signature_b64":"r6IfWhQEyKmtFVq2FgZzQ4Mbv3rpBqj6EOxuzYnaW6W4/bLTwVLsOjgx+untERM9V8iXBI5OcP1Q6ZBCn6tJDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bd39dd6c7abdded041ba097a8d39e8e3d2d9e8913054315c4206cbf75671debd","last_reissued_at":"2026-07-05T10:56:41.937414Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:56:41.937414Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Fine-Tuning LLMs for Low-Resource Dialect Translation: The Case of Lebanese","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ali Chehab, Silvana Yakhni","submitted_at":"2025-04-30T18:33:53Z","abstract_excerpt":"This paper examines the effectiveness of Large Language Models (LLMs) in translating the low-resource Lebanese dialect, focusing on the impact of culturally authentic data versus larger translated datasets. We compare three fine-tuning approaches: Basic, contrastive, and grammar-hint tuning, using open-source Aya23 models. Experiments reveal that models fine-tuned on a smaller but culturally aware Lebanese dataset (LW) consistently outperform those trained on larger, non-native data. The best results were achieved through contrastive fine-tuning paired with contrastive prompting, which indicat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.00114","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/2505.00114/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":"2505.00114","created_at":"2026-07-05T10:56:41.937469+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.00114v1","created_at":"2026-07-05T10:56:41.937469+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.00114","created_at":"2026-07-05T10:56:41.937469+00:00"},{"alias_kind":"pith_short_12","alias_value":"XU4523D2XXPN","created_at":"2026-07-05T10:56:41.937469+00:00"},{"alias_kind":"pith_short_16","alias_value":"XU4523D2XXPNAQN2","created_at":"2026-07-05T10:56:41.937469+00:00"},{"alias_kind":"pith_short_8","alias_value":"XU4523D2","created_at":"2026-07-05T10:56:41.937469+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/XU4523D2XXPNAQN2BF5I2OPI4P","json":"https://pith.science/pith/XU4523D2XXPNAQN2BF5I2OPI4P.json","graph_json":"https://pith.science/api/pith-number/XU4523D2XXPNAQN2BF5I2OPI4P/graph.json","events_json":"https://pith.science/api/pith-number/XU4523D2XXPNAQN2BF5I2OPI4P/events.json","paper":"https://pith.science/paper/XU4523D2"},"agent_actions":{"view_html":"https://pith.science/pith/XU4523D2XXPNAQN2BF5I2OPI4P","download_json":"https://pith.science/pith/XU4523D2XXPNAQN2BF5I2OPI4P.json","view_paper":"https://pith.science/paper/XU4523D2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.00114&json=true","fetch_graph":"https://pith.science/api/pith-number/XU4523D2XXPNAQN2BF5I2OPI4P/graph.json","fetch_events":"https://pith.science/api/pith-number/XU4523D2XXPNAQN2BF5I2OPI4P/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XU4523D2XXPNAQN2BF5I2OPI4P/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XU4523D2XXPNAQN2BF5I2OPI4P/action/storage_attestation","attest_author":"https://pith.science/pith/XU4523D2XXPNAQN2BF5I2OPI4P/action/author_attestation","sign_citation":"https://pith.science/pith/XU4523D2XXPNAQN2BF5I2OPI4P/action/citation_signature","submit_replication":"https://pith.science/pith/XU4523D2XXPNAQN2BF5I2OPI4P/action/replication_record"}},"created_at":"2026-07-05T10:56:41.937469+00:00","updated_at":"2026-07-05T10:56:41.937469+00:00"}