{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:VQCA5572YI3T26C27PEUTHQJUT","short_pith_number":"pith:VQCA5572","schema_version":"1.0","canonical_sha256":"ac040ef7fac2373d785afbc9499e09a4c0a8d42a3c275f19c40a4df561e00228","source":{"kind":"arxiv","id":"2504.09753","version":3},"attestation_state":"computed","paper":{"title":"Improving Multilingual Capabilities with Cultural and Local Knowledge in Large Language Models While Enhancing Native Performance","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Drishti Sharma, Hamza Farooq, Jebish Purbey, Kanwal Mehreen, Muhammad Arham, Ram Mohan Rao Kadiyala, Siddartha Pullakhandam, Siddhant Gupta, Suman Debnath","submitted_at":"2025-04-13T23:10:13Z","abstract_excerpt":"Large Language Models (LLMs) have shown remarkable capabilities, but their development has primarily focused on English and other high-resource languages, leaving many languages underserved. We present our latest Hindi-English bi-lingual LLM \\textbf{Mantra-14B} with ~3\\% average improvement in benchmark scores over both languages, outperforming models twice its size. Using a curated dataset composed of English and Hindi instruction data of 485K samples, we instruction tuned models such as Qwen-2.5-14B-Instruct and Phi-4 to improve performance over both English and Hindi. Our experiments encomp"},"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":"2504.09753","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-13T23:10:13Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"bb5657e99adb29d2cb7e42f6093b9da100e4c090d7c18c298cd2e699234a670f","abstract_canon_sha256":"b92b543d364c724a6715bfce8e0e7b6399752e68cebd1aafc552d5335a297122"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:45:57.281736Z","signature_b64":"yaFPSSEMEyLOvfQFd3mRj0pHaM8eeye7VpK4dKuwsJ4K/fYLnJDAOx1RbU/fMWlqTr6e3uot4TrBv51EsMYMAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ac040ef7fac2373d785afbc9499e09a4c0a8d42a3c275f19c40a4df561e00228","last_reissued_at":"2026-07-05T11:45:57.281203Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:45:57.281203Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Improving Multilingual Capabilities with Cultural and Local Knowledge in Large Language Models While Enhancing Native Performance","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Drishti Sharma, Hamza Farooq, Jebish Purbey, Kanwal Mehreen, Muhammad Arham, Ram Mohan Rao Kadiyala, Siddartha Pullakhandam, Siddhant Gupta, Suman Debnath","submitted_at":"2025-04-13T23:10:13Z","abstract_excerpt":"Large Language Models (LLMs) have shown remarkable capabilities, but their development has primarily focused on English and other high-resource languages, leaving many languages underserved. We present our latest Hindi-English bi-lingual LLM \\textbf{Mantra-14B} with ~3\\% average improvement in benchmark scores over both languages, outperforming models twice its size. Using a curated dataset composed of English and Hindi instruction data of 485K samples, we instruction tuned models such as Qwen-2.5-14B-Instruct and Phi-4 to improve performance over both English and Hindi. Our experiments encomp"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.09753","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/2504.09753/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":"2504.09753","created_at":"2026-07-05T11:45:57.281267+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.09753v3","created_at":"2026-07-05T11:45:57.281267+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.09753","created_at":"2026-07-05T11:45:57.281267+00:00"},{"alias_kind":"pith_short_12","alias_value":"VQCA5572YI3T","created_at":"2026-07-05T11:45:57.281267+00:00"},{"alias_kind":"pith_short_16","alias_value":"VQCA5572YI3T26C2","created_at":"2026-07-05T11:45:57.281267+00:00"},{"alias_kind":"pith_short_8","alias_value":"VQCA5572","created_at":"2026-07-05T11:45:57.281267+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/VQCA5572YI3T26C27PEUTHQJUT","json":"https://pith.science/pith/VQCA5572YI3T26C27PEUTHQJUT.json","graph_json":"https://pith.science/api/pith-number/VQCA5572YI3T26C27PEUTHQJUT/graph.json","events_json":"https://pith.science/api/pith-number/VQCA5572YI3T26C27PEUTHQJUT/events.json","paper":"https://pith.science/paper/VQCA5572"},"agent_actions":{"view_html":"https://pith.science/pith/VQCA5572YI3T26C27PEUTHQJUT","download_json":"https://pith.science/pith/VQCA5572YI3T26C27PEUTHQJUT.json","view_paper":"https://pith.science/paper/VQCA5572","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.09753&json=true","fetch_graph":"https://pith.science/api/pith-number/VQCA5572YI3T26C27PEUTHQJUT/graph.json","fetch_events":"https://pith.science/api/pith-number/VQCA5572YI3T26C27PEUTHQJUT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VQCA5572YI3T26C27PEUTHQJUT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VQCA5572YI3T26C27PEUTHQJUT/action/storage_attestation","attest_author":"https://pith.science/pith/VQCA5572YI3T26C27PEUTHQJUT/action/author_attestation","sign_citation":"https://pith.science/pith/VQCA5572YI3T26C27PEUTHQJUT/action/citation_signature","submit_replication":"https://pith.science/pith/VQCA5572YI3T26C27PEUTHQJUT/action/replication_record"}},"created_at":"2026-07-05T11:45:57.281267+00:00","updated_at":"2026-07-05T11:45:57.281267+00:00"}