{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UN35WBMAFWUW4NRTLHUKJAY72W","short_pith_number":"pith:UN35WBMA","schema_version":"1.0","canonical_sha256":"a377db05802da96e363359e8a4831fd59a6899b781bb93f8a3a5e29a073b072f","source":{"kind":"arxiv","id":"2608.13580","version":1},"attestation_state":"computed","paper":{"title":"Jais 2: A Family of Arabic-Centric Open Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Aaryamonvikram Singh, Abdelaziz Bounhar, AbdelRahman Elbadawy, Abdelrahman Sadallah, Abed Alhakim Freihat, Abhishek Maiti, Ahmad Chamma, Ahmed Frikha, Ali El Filali, Ali Mekky, Amr Mohamed, Avraham Sheinin, Awantika Shukla, Biswajit Mishra, Dani Bouch, Daniil Orel, Etienne Goffinet, Evan Dufraisse, Fajri Koto, Farah Atif, George Ibrahim, Gokulakrishnan Ramakrishnan, Guokan Shang, Gurpreet Gosal, Hadi Abdine, Haonan Li, Hasan Iqbal, Hector Xuguang Ren, Joel Hestness, Kareem Elozeiri, Kareem Elzeky, Larry Murray, Mervat Abassy, Michalis Vazirgiannis, Mohamed Anwar, Momina Ahsan, Mostafa Awad, Natalia Vassilieva, Neha Sengupta, Nurdaulet Mukhituly, Omar El Herraoui, Onkar Pandit, Parvez Mullah, Preslav Nakov, Rahul Pal, Rania Elbadry, Rituraj Joshi, Saadeldine Eletter, Sajid siddiki, Samta Kamboj, Samujjwal Ghosh, Sarah Albarri, Sarath Chandran, Sarfraz Ahmad, Sunil Kumar Sahu, Xudong Han, Yuxia Wang, Zainul Abedien Ahmed Quraishi, Zhengzhong Liu, Zhuohan Xie","submitted_at":"2026-07-07T13:54:49Z","abstract_excerpt":"Jais 2 is a family of Arabic-centric large language models developed jointly by MBZUAI, Cerebras, and Inception, designed to advance Arabic-centric language modeling, with strong performance across the Arabic and culturally grounded benchmarks evaluated in this report. The family includes, to our knowledge, the largest open Arabic-centric LLM trained from scratch at 70B parameters, and a competitive 8B-parameter variant among the evaluated open models. A custom Arabic-centric vocabulary enables efficient training and inference. In addition, an optimized architecture and training recipe yield h"},"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":"2608.13580","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-07T13:54:49Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"09b531e3b46b27324f3f483192593274d685be7c13dc3e459c3958ff7abc58cd","abstract_canon_sha256":"b4ed50776bcb5449327b29b008fa6b14920267892a66b5f391bef0b2371fc19a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-17T00:08:57.194489Z","signature_b64":"vF20DfejTj90uNN5x/21qWvdAo7Ln/eCtHIJL3JdVzuJ0ErOzJmrOkPUbdmApjVOF4L2vtHQxYijT7viyHjnCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a377db05802da96e363359e8a4831fd59a6899b781bb93f8a3a5e29a073b072f","last_reissued_at":"2026-08-17T00:08:57.192877Z","signature_status":"signed_v1","first_computed_at":"2026-08-17T00:08:57.192877Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Jais 2: A Family of Arabic-Centric Open Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Aaryamonvikram Singh, Abdelaziz Bounhar, AbdelRahman Elbadawy, Abdelrahman Sadallah, Abed Alhakim Freihat, Abhishek Maiti, Ahmad Chamma, Ahmed Frikha, Ali El Filali, Ali Mekky, Amr Mohamed, Avraham Sheinin, Awantika Shukla, Biswajit Mishra, Dani Bouch, Daniil Orel, Etienne Goffinet, Evan Dufraisse, Fajri Koto, Farah Atif, George Ibrahim, Gokulakrishnan Ramakrishnan, Guokan Shang, Gurpreet Gosal, Hadi Abdine, Haonan Li, Hasan Iqbal, Hector Xuguang Ren, Joel Hestness, Kareem Elozeiri, Kareem Elzeky, Larry Murray, Mervat Abassy, Michalis Vazirgiannis, Mohamed Anwar, Momina Ahsan, Mostafa Awad, Natalia Vassilieva, Neha Sengupta, Nurdaulet Mukhituly, Omar El Herraoui, Onkar Pandit, Parvez Mullah, Preslav Nakov, Rahul Pal, Rania Elbadry, Rituraj Joshi, Saadeldine Eletter, Sajid siddiki, Samta Kamboj, Samujjwal Ghosh, Sarah Albarri, Sarath Chandran, Sarfraz Ahmad, Sunil Kumar Sahu, Xudong Han, Yuxia Wang, Zainul Abedien Ahmed Quraishi, Zhengzhong Liu, Zhuohan Xie","submitted_at":"2026-07-07T13:54:49Z","abstract_excerpt":"Jais 2 is a family of Arabic-centric large language models developed jointly by MBZUAI, Cerebras, and Inception, designed to advance Arabic-centric language modeling, with strong performance across the Arabic and culturally grounded benchmarks evaluated in this report. The family includes, to our knowledge, the largest open Arabic-centric LLM trained from scratch at 70B parameters, and a competitive 8B-parameter variant among the evaluated open models. A custom Arabic-centric vocabulary enables efficient training and inference. In addition, an optimized architecture and training recipe yield h"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.13580","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/2608.13580/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":"2608.13580","created_at":"2026-08-17T00:08:57.192762+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.13580v1","created_at":"2026-08-17T00:08:57.192762+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.13580","created_at":"2026-08-17T00:08:57.192762+00:00"},{"alias_kind":"pith_short_12","alias_value":"UN35WBMAFWUW","created_at":"2026-08-17T00:08:57.192762+00:00"},{"alias_kind":"pith_short_16","alias_value":"UN35WBMAFWUW4NRT","created_at":"2026-08-17T00:08:57.192762+00:00"},{"alias_kind":"pith_short_8","alias_value":"UN35WBMA","created_at":"2026-08-17T00:08:57.192762+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/UN35WBMAFWUW4NRTLHUKJAY72W","json":"https://pith.science/pith/UN35WBMAFWUW4NRTLHUKJAY72W.json","graph_json":"https://pith.science/api/pith-number/UN35WBMAFWUW4NRTLHUKJAY72W/graph.json","events_json":"https://pith.science/api/pith-number/UN35WBMAFWUW4NRTLHUKJAY72W/events.json","paper":"https://pith.science/paper/UN35WBMA"},"agent_actions":{"view_html":"https://pith.science/pith/UN35WBMAFWUW4NRTLHUKJAY72W","download_json":"https://pith.science/pith/UN35WBMAFWUW4NRTLHUKJAY72W.json","view_paper":"https://pith.science/paper/UN35WBMA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.13580&json=true","fetch_graph":"https://pith.science/api/pith-number/UN35WBMAFWUW4NRTLHUKJAY72W/graph.json","fetch_events":"https://pith.science/api/pith-number/UN35WBMAFWUW4NRTLHUKJAY72W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UN35WBMAFWUW4NRTLHUKJAY72W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UN35WBMAFWUW4NRTLHUKJAY72W/action/storage_attestation","attest_author":"https://pith.science/pith/UN35WBMAFWUW4NRTLHUKJAY72W/action/author_attestation","sign_citation":"https://pith.science/pith/UN35WBMAFWUW4NRTLHUKJAY72W/action/citation_signature","submit_replication":"https://pith.science/pith/UN35WBMAFWUW4NRTLHUKJAY72W/action/replication_record"}},"created_at":"2026-08-17T00:08:57.192762+00:00","updated_at":"2026-08-17T00:08:57.192762+00:00"}