{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:AOQBCISK3S64FCBRBCAHQ5VPHC","short_pith_number":"pith:AOQBCISK","schema_version":"1.0","canonical_sha256":"03a011224adcbdc2883108807876af389a5a64f2721b8ef40c31ec26147e83a5","source":{"kind":"arxiv","id":"2505.22293","version":1},"attestation_state":"computed","paper":{"title":"Compensating for Data with Reasoning: Low-Resource Machine Translation with LLMs","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Samuel Frontull, Thomas Str\\\"ohle","submitted_at":"2025-05-28T12:29:05Z","abstract_excerpt":"Large Language Models (LLMs) have demonstrated strong capabilities in multilingual machine translation, sometimes even outperforming traditional neural systems. However, previous research has highlighted the challenges of using LLMs, particularly with prompt engineering, for low-resource languages. In this work, we introduce Fragment-Shot Prompting, a novel in-context learning method that segments input and retrieves translation examples based on syntactic coverage, along with Pivoted Fragment-Shot, an extension that enables translation without direct parallel data. We evaluate these methods u"},"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":"2505.22293","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-28T12:29:05Z","cross_cats_sorted":[],"title_canon_sha256":"ddcfca575fb45f696fb2f1a49e2c5f818dd313479456e0f8af8c15762a9996fb","abstract_canon_sha256":"9d2ecb819e2b4374ebf69b3e289cc68c92712ced6232fc8211e0f70686530dd4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:12.664500Z","signature_b64":"Bdd0hCRoEjAlAoB9VpKrDXDYAuoa3pvaEK/3jRzVJhJv1v5gbc5AMdweo6kkXeP2qLO+CdJQNxZ8VHn6oxYJCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"03a011224adcbdc2883108807876af389a5a64f2721b8ef40c31ec26147e83a5","last_reissued_at":"2026-07-05T11:11:12.663993Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:12.663993Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Compensating for Data with Reasoning: Low-Resource Machine Translation with LLMs","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Samuel Frontull, Thomas Str\\\"ohle","submitted_at":"2025-05-28T12:29:05Z","abstract_excerpt":"Large Language Models (LLMs) have demonstrated strong capabilities in multilingual machine translation, sometimes even outperforming traditional neural systems. However, previous research has highlighted the challenges of using LLMs, particularly with prompt engineering, for low-resource languages. In this work, we introduce Fragment-Shot Prompting, a novel in-context learning method that segments input and retrieves translation examples based on syntactic coverage, along with Pivoted Fragment-Shot, an extension that enables translation without direct parallel data. We evaluate these methods u"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.22293","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/2505.22293/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":"2505.22293","created_at":"2026-07-05T11:11:12.664054+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.22293v1","created_at":"2026-07-05T11:11:12.664054+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.22293","created_at":"2026-07-05T11:11:12.664054+00:00"},{"alias_kind":"pith_short_12","alias_value":"AOQBCISK3S64","created_at":"2026-07-05T11:11:12.664054+00:00"},{"alias_kind":"pith_short_16","alias_value":"AOQBCISK3S64FCBR","created_at":"2026-07-05T11:11:12.664054+00:00"},{"alias_kind":"pith_short_8","alias_value":"AOQBCISK","created_at":"2026-07-05T11:11:12.664054+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.06420","citing_title":"A Komi-Yazva--Russian Parallel Corpus and Evaluation Protocol for Zero- and Few-Shot LLM Translation","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2504.01919","citing_title":"Bridging the Linguistic Divide: A Survey on Leveraging Large Language Models for Machine Translation","ref_index":78,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18758","citing_title":"Syntax as a Rosetta Stone: Universal Dependencies for In-Context Coptic Translation","ref_index":39,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AOQBCISK3S64FCBRBCAHQ5VPHC","json":"https://pith.science/pith/AOQBCISK3S64FCBRBCAHQ5VPHC.json","graph_json":"https://pith.science/api/pith-number/AOQBCISK3S64FCBRBCAHQ5VPHC/graph.json","events_json":"https://pith.science/api/pith-number/AOQBCISK3S64FCBRBCAHQ5VPHC/events.json","paper":"https://pith.science/paper/AOQBCISK"},"agent_actions":{"view_html":"https://pith.science/pith/AOQBCISK3S64FCBRBCAHQ5VPHC","download_json":"https://pith.science/pith/AOQBCISK3S64FCBRBCAHQ5VPHC.json","view_paper":"https://pith.science/paper/AOQBCISK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.22293&json=true","fetch_graph":"https://pith.science/api/pith-number/AOQBCISK3S64FCBRBCAHQ5VPHC/graph.json","fetch_events":"https://pith.science/api/pith-number/AOQBCISK3S64FCBRBCAHQ5VPHC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AOQBCISK3S64FCBRBCAHQ5VPHC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AOQBCISK3S64FCBRBCAHQ5VPHC/action/storage_attestation","attest_author":"https://pith.science/pith/AOQBCISK3S64FCBRBCAHQ5VPHC/action/author_attestation","sign_citation":"https://pith.science/pith/AOQBCISK3S64FCBRBCAHQ5VPHC/action/citation_signature","submit_replication":"https://pith.science/pith/AOQBCISK3S64FCBRBCAHQ5VPHC/action/replication_record"}},"created_at":"2026-07-05T11:11:12.664054+00:00","updated_at":"2026-07-05T11:11:12.664054+00:00"}