{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:P7XG6DS2TWB23RWOG3IGARW3VM","short_pith_number":"pith:P7XG6DS2","schema_version":"1.0","canonical_sha256":"7fee6f0e5a9d83adc6ce36d06046dbab3294ce6e38b99e817ee81e6b1f0891b8","source":{"kind":"arxiv","id":"2012.15520","version":2},"attestation_state":"computed","paper":{"title":"AraGPT2: Pre-Trained Transformer for Arabic Language Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Fady Baly, Hazem Hajj, Wissam Antoun","submitted_at":"2020-12-31T09:48:05Z","abstract_excerpt":"Recently, pre-trained transformer-based architectures have proven to be very efficient at language modeling and understanding, given that they are trained on a large enough corpus. Applications in language generation for Arabic are still lagging in comparison to other NLP advances primarily due to the lack of advanced Arabic language generation models. In this paper, we develop the first advanced Arabic language generation model, AraGPT2, trained from scratch on a large Arabic corpus of internet text and news articles. Our largest model, AraGPT2-mega, has 1.46 billion parameters, which makes i"},"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":"2012.15520","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-12-31T09:48:05Z","cross_cats_sorted":[],"title_canon_sha256":"e7758e806291f07f3d1c086ef9973fc48d0b537519030dd5158a5cbfcb505a83","abstract_canon_sha256":"ddec160d18ea5c8ced02461af1fd5e3e3a434742a2ee7f70f9c5c2f5563ee63f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:21:00.240443Z","signature_b64":"JMQ/+3mIS1rNW/5Hx4WvvBXF0+OSJxEd4GKssUN60j/HjMtCPohO1GjU6q98M0zO/oX8ZRpor5INAUi+6MNxDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7fee6f0e5a9d83adc6ce36d06046dbab3294ce6e38b99e817ee81e6b1f0891b8","last_reissued_at":"2026-07-05T02:21:00.239946Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:21:00.239946Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AraGPT2: Pre-Trained Transformer for Arabic Language Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Fady Baly, Hazem Hajj, Wissam Antoun","submitted_at":"2020-12-31T09:48:05Z","abstract_excerpt":"Recently, pre-trained transformer-based architectures have proven to be very efficient at language modeling and understanding, given that they are trained on a large enough corpus. Applications in language generation for Arabic are still lagging in comparison to other NLP advances primarily due to the lack of advanced Arabic language generation models. In this paper, we develop the first advanced Arabic language generation model, AraGPT2, trained from scratch on a large Arabic corpus of internet text and news articles. Our largest model, AraGPT2-mega, has 1.46 billion parameters, which makes i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2012.15520","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/2012.15520/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":"2012.15520","created_at":"2026-07-05T02:21:00.240005+00:00"},{"alias_kind":"arxiv_version","alias_value":"2012.15520v2","created_at":"2026-07-05T02:21:00.240005+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2012.15520","created_at":"2026-07-05T02:21:00.240005+00:00"},{"alias_kind":"pith_short_12","alias_value":"P7XG6DS2TWB2","created_at":"2026-07-05T02:21:00.240005+00:00"},{"alias_kind":"pith_short_16","alias_value":"P7XG6DS2TWB23RWO","created_at":"2026-07-05T02:21:00.240005+00:00"},{"alias_kind":"pith_short_8","alias_value":"P7XG6DS2","created_at":"2026-07-05T02:21:00.240005+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.21108","citing_title":"Machine learning and emoji prediction: How much accuracy can MARBERT achieve?","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/P7XG6DS2TWB23RWOG3IGARW3VM","json":"https://pith.science/pith/P7XG6DS2TWB23RWOG3IGARW3VM.json","graph_json":"https://pith.science/api/pith-number/P7XG6DS2TWB23RWOG3IGARW3VM/graph.json","events_json":"https://pith.science/api/pith-number/P7XG6DS2TWB23RWOG3IGARW3VM/events.json","paper":"https://pith.science/paper/P7XG6DS2"},"agent_actions":{"view_html":"https://pith.science/pith/P7XG6DS2TWB23RWOG3IGARW3VM","download_json":"https://pith.science/pith/P7XG6DS2TWB23RWOG3IGARW3VM.json","view_paper":"https://pith.science/paper/P7XG6DS2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2012.15520&json=true","fetch_graph":"https://pith.science/api/pith-number/P7XG6DS2TWB23RWOG3IGARW3VM/graph.json","fetch_events":"https://pith.science/api/pith-number/P7XG6DS2TWB23RWOG3IGARW3VM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/P7XG6DS2TWB23RWOG3IGARW3VM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/P7XG6DS2TWB23RWOG3IGARW3VM/action/storage_attestation","attest_author":"https://pith.science/pith/P7XG6DS2TWB23RWOG3IGARW3VM/action/author_attestation","sign_citation":"https://pith.science/pith/P7XG6DS2TWB23RWOG3IGARW3VM/action/citation_signature","submit_replication":"https://pith.science/pith/P7XG6DS2TWB23RWOG3IGARW3VM/action/replication_record"}},"created_at":"2026-07-05T02:21:00.240005+00:00","updated_at":"2026-07-05T02:21:00.240005+00:00"}