{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:K2WVAVFYHETJRPFX2EZP5AZ6DQ","short_pith_number":"pith:K2WVAVFY","schema_version":"1.0","canonical_sha256":"56ad5054b8392698bcb7d132fe833e1c38c0d42ee9ca3e0ef935792af4ac307c","source":{"kind":"arxiv","id":"2402.17733","version":1},"attestation_state":"computed","paper":{"title":"Tower: An Open Multilingual Large Language Model for Translation-Related Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Amin Farajian, Andr\\'e F.T. Martins, Ben Peters, Duarte M. Alves, Jo\\~ao Alves, Jos\\'e G.C. de Souza, Jos\\'e Pombal, Nuno M. Guerreiro, Patrick Fernandes, Pedro H. Martins, Pierre Colombo, Ricardo Rei, Sweta Agrawal","submitted_at":"2024-02-27T18:09:36Z","abstract_excerpt":"While general-purpose large language models (LLMs) demonstrate proficiency on multiple tasks within the domain of translation, approaches based on open LLMs are competitive only when specializing on a single task. In this paper, we propose a recipe for tailoring LLMs to multiple tasks present in translation workflows. We perform continued pretraining on a multilingual mixture of monolingual and parallel data, creating TowerBase, followed by finetuning on instructions relevant for translation processes, creating TowerInstruct. Our final model surpasses open alternatives on several tasks relevan"},"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":"2402.17733","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-27T18:09:36Z","cross_cats_sorted":[],"title_canon_sha256":"060fa128e18ac5f0436c69f6f36b942695c4ba0eba7f32f7e11ef8cc6602e644","abstract_canon_sha256":"41e2b25ccc8daa73a29f80c2f9a220bbf6353060ffc625b7095cc4413be5c693"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:49:56.871191Z","signature_b64":"7Zp7/JYC/K80+P4ttBpqlrcBUUo8vEV8+VXbV5LYSaHaSMXLOc/3bivJLfEzvUKzKSn2JEtf1ebzv3MM8K+QDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"56ad5054b8392698bcb7d132fe833e1c38c0d42ee9ca3e0ef935792af4ac307c","last_reissued_at":"2026-07-05T07:49:56.870703Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:49:56.870703Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Tower: An Open Multilingual Large Language Model for Translation-Related Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Amin Farajian, Andr\\'e F.T. Martins, Ben Peters, Duarte M. Alves, Jo\\~ao Alves, Jos\\'e G.C. de Souza, Jos\\'e Pombal, Nuno M. Guerreiro, Patrick Fernandes, Pedro H. Martins, Pierre Colombo, Ricardo Rei, Sweta Agrawal","submitted_at":"2024-02-27T18:09:36Z","abstract_excerpt":"While general-purpose large language models (LLMs) demonstrate proficiency on multiple tasks within the domain of translation, approaches based on open LLMs are competitive only when specializing on a single task. In this paper, we propose a recipe for tailoring LLMs to multiple tasks present in translation workflows. We perform continued pretraining on a multilingual mixture of monolingual and parallel data, creating TowerBase, followed by finetuning on instructions relevant for translation processes, creating TowerInstruct. Our final model surpasses open alternatives on several tasks relevan"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.17733","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/2402.17733/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":"2402.17733","created_at":"2026-07-05T07:49:56.870764+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.17733v1","created_at":"2026-07-05T07:49:56.870764+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.17733","created_at":"2026-07-05T07:49:56.870764+00:00"},{"alias_kind":"pith_short_12","alias_value":"K2WVAVFYHETJ","created_at":"2026-07-05T07:49:56.870764+00:00"},{"alias_kind":"pith_short_16","alias_value":"K2WVAVFYHETJRPFX","created_at":"2026-07-05T07:49:56.870764+00:00"},{"alias_kind":"pith_short_8","alias_value":"K2WVAVFY","created_at":"2026-07-05T07:49:56.870764+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.04730","citing_title":"Multilingual Long-Form Speech Instruction Following: KIT's Submission to IWSLT 2026","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2604.22520","citing_title":"RouteLMT: Learned Sample Routing for Hybrid LLM Translation Deployment","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04704","citing_title":"UVMarvel: an Automated LLM-aided UVM Machine for Subsystem-level RTL Verification","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01580","citing_title":"Model Merging: Foundations and Algorithms","ref_index":5,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/K2WVAVFYHETJRPFX2EZP5AZ6DQ","json":"https://pith.science/pith/K2WVAVFYHETJRPFX2EZP5AZ6DQ.json","graph_json":"https://pith.science/api/pith-number/K2WVAVFYHETJRPFX2EZP5AZ6DQ/graph.json","events_json":"https://pith.science/api/pith-number/K2WVAVFYHETJRPFX2EZP5AZ6DQ/events.json","paper":"https://pith.science/paper/K2WVAVFY"},"agent_actions":{"view_html":"https://pith.science/pith/K2WVAVFYHETJRPFX2EZP5AZ6DQ","download_json":"https://pith.science/pith/K2WVAVFYHETJRPFX2EZP5AZ6DQ.json","view_paper":"https://pith.science/paper/K2WVAVFY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.17733&json=true","fetch_graph":"https://pith.science/api/pith-number/K2WVAVFYHETJRPFX2EZP5AZ6DQ/graph.json","fetch_events":"https://pith.science/api/pith-number/K2WVAVFYHETJRPFX2EZP5AZ6DQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/K2WVAVFYHETJRPFX2EZP5AZ6DQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/K2WVAVFYHETJRPFX2EZP5AZ6DQ/action/storage_attestation","attest_author":"https://pith.science/pith/K2WVAVFYHETJRPFX2EZP5AZ6DQ/action/author_attestation","sign_citation":"https://pith.science/pith/K2WVAVFYHETJRPFX2EZP5AZ6DQ/action/citation_signature","submit_replication":"https://pith.science/pith/K2WVAVFYHETJRPFX2EZP5AZ6DQ/action/replication_record"}},"created_at":"2026-07-05T07:49:56.870764+00:00","updated_at":"2026-07-05T07:49:56.870764+00:00"}