{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6KQJR5TYD244WBIHDPW4RZJBAZ","short_pith_number":"pith:6KQJR5TY","schema_version":"1.0","canonical_sha256":"f2a098f6781eb9cb05071bedc8e521067cf5dc90ece3c359750bf04bd332d677","source":{"kind":"arxiv","id":"2401.13303","version":2},"attestation_state":"computed","paper":{"title":"MaLA-500: Massive Language Adaptation of Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Andr\\'e F. T. Martins, Hinrich Sch\\\"utze, J\\\"org Tiedemann, Peiqin Lin, Shaoxiong Ji","submitted_at":"2024-01-24T08:57:39Z","abstract_excerpt":"Large language models (LLMs) have advanced the state of the art in natural language processing. However, their predominant design for English or a limited set of languages creates a substantial gap in their effectiveness for low-resource languages. To bridge this gap, we introduce MaLA-500, a novel large language model designed to cover an extensive range of 534 languages. To train MaLA-500, we employ vocabulary extension and continued pretraining on LLaMA 2 with Glot500-c. Our intrinsic evaluation demonstrates that MaLA-500 is better at predicting the given texts of low-resource languages tha"},"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":"2401.13303","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-24T08:57:39Z","cross_cats_sorted":[],"title_canon_sha256":"a852654eeafded45e773190ad5934c9814010c612ee93ddbdbc791ea8e62e039","abstract_canon_sha256":"8c53534e7bc4e6fc91d03652944aedaf47e809fae2c94686e977479463913a18"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:03:48.398158Z","signature_b64":"MQwP532jOR60U8YG47ejBzxvea94qpFLyop+lnPePE/NPDuAUZ8cboKsDqkHR9zNYcREcrXhIJLWjM31r0w7Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f2a098f6781eb9cb05071bedc8e521067cf5dc90ece3c359750bf04bd332d677","last_reissued_at":"2026-07-05T08:03:48.397739Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:03:48.397739Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MaLA-500: Massive Language Adaptation of Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Andr\\'e F. T. Martins, Hinrich Sch\\\"utze, J\\\"org Tiedemann, Peiqin Lin, Shaoxiong Ji","submitted_at":"2024-01-24T08:57:39Z","abstract_excerpt":"Large language models (LLMs) have advanced the state of the art in natural language processing. However, their predominant design for English or a limited set of languages creates a substantial gap in their effectiveness for low-resource languages. To bridge this gap, we introduce MaLA-500, a novel large language model designed to cover an extensive range of 534 languages. To train MaLA-500, we employ vocabulary extension and continued pretraining on LLaMA 2 with Glot500-c. Our intrinsic evaluation demonstrates that MaLA-500 is better at predicting the given texts of low-resource languages tha"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.13303","kind":"arxiv","version":2},"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/2401.13303/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":"2401.13303","created_at":"2026-07-05T08:03:48.397796+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.13303v2","created_at":"2026-07-05T08:03:48.397796+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.13303","created_at":"2026-07-05T08:03:48.397796+00:00"},{"alias_kind":"pith_short_12","alias_value":"6KQJR5TYD244","created_at":"2026-07-05T08:03:48.397796+00:00"},{"alias_kind":"pith_short_16","alias_value":"6KQJR5TYD244WBIH","created_at":"2026-07-05T08:03:48.397796+00:00"},{"alias_kind":"pith_short_8","alias_value":"6KQJR5TY","created_at":"2026-07-05T08:03:48.397796+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.06738","citing_title":"Modular Monolingual Adaptation using Pretrained Language Models","ref_index":66,"is_internal_anchor":false},{"citing_arxiv_id":"2412.12686","citing_title":"Exploring Cross-lingual Latent Transplantation: Mutual Opportunities and Open Challenges","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13225","citing_title":"Mix, Don't Tune: Bilingual Pre-Training Outperforms Hyperparameter Search in Data-Constrained Settings","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06115","citing_title":"CrossCult-KIBench: A Benchmark for Cross-Cultural Knowledge Insertion in MLLMs","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06115","citing_title":"CrossCult-KIBench: A Benchmark for Cross-Cultural Knowledge Insertion in MLLMs","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6KQJR5TYD244WBIHDPW4RZJBAZ","json":"https://pith.science/pith/6KQJR5TYD244WBIHDPW4RZJBAZ.json","graph_json":"https://pith.science/api/pith-number/6KQJR5TYD244WBIHDPW4RZJBAZ/graph.json","events_json":"https://pith.science/api/pith-number/6KQJR5TYD244WBIHDPW4RZJBAZ/events.json","paper":"https://pith.science/paper/6KQJR5TY"},"agent_actions":{"view_html":"https://pith.science/pith/6KQJR5TYD244WBIHDPW4RZJBAZ","download_json":"https://pith.science/pith/6KQJR5TYD244WBIHDPW4RZJBAZ.json","view_paper":"https://pith.science/paper/6KQJR5TY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.13303&json=true","fetch_graph":"https://pith.science/api/pith-number/6KQJR5TYD244WBIHDPW4RZJBAZ/graph.json","fetch_events":"https://pith.science/api/pith-number/6KQJR5TYD244WBIHDPW4RZJBAZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6KQJR5TYD244WBIHDPW4RZJBAZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6KQJR5TYD244WBIHDPW4RZJBAZ/action/storage_attestation","attest_author":"https://pith.science/pith/6KQJR5TYD244WBIHDPW4RZJBAZ/action/author_attestation","sign_citation":"https://pith.science/pith/6KQJR5TYD244WBIHDPW4RZJBAZ/action/citation_signature","submit_replication":"https://pith.science/pith/6KQJR5TYD244WBIHDPW4RZJBAZ/action/replication_record"}},"created_at":"2026-07-05T08:03:48.397796+00:00","updated_at":"2026-07-05T08:03:48.397796+00:00"}