{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:EGCE3ZTISSAB7QJF7FVL4ODTZ4","short_pith_number":"pith:EGCE3ZTI","schema_version":"1.0","canonical_sha256":"21844de66894801fc125f96abe3873cf123cf05991103513c85fbb605a8c4fda","source":{"kind":"arxiv","id":"2407.14885","version":1},"attestation_state":"computed","paper":{"title":"Falcon2-11B Technical Report","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.CL","authors_text":"Ankit Singh, Basma El Amel Boussaha, Giulia Campesan, Hakim Hacid, Hamza Alobeidli, Kirill Fedyanin, Leen Al Qadi, Maksim Velikanov, Mohamed El Amine Seddik, Mohammed Al-Yafeai, Mugariya Farooq, Nilabhra Roy Chowdhury, Quentin Malartic, Reda Alami, Ruxandra Cojocaru, Sanath Narayan, Yasser Abdelaziz Dahou Djilali","submitted_at":"2024-07-20T14:23:15Z","abstract_excerpt":"We introduce Falcon2-11B, a foundation model trained on over five trillion tokens, and its multimodal counterpart, Falcon2-11B-vlm, which is a vision-to-text model. We report our findings during the training of the Falcon2-11B which follows a multi-stage approach where the early stages are distinguished by their context length and a final stage where we use a curated, high-quality dataset. Additionally, we report the effect of doubling the batch size mid-training and how training loss spikes are affected by the learning rate. The downstream performance of the foundation model is evaluated on e"},"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":"2407.14885","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-07-20T14:23:15Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"16ba1630217002ddcbabf47eeb7992f0dac7a8923175003b5c7af80be2b28fd2","abstract_canon_sha256":"d1889510fb537ed9a668a79987fbab8331d05d6eb6357fb75ad0a98ee5f0bac2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:46:35.582690Z","signature_b64":"3fRjZ8BnRO/gdsCRzUAVXRkEh0sHqbhuuD7a6lzo2jMbBIpxkeTtbJi1dL8uZ/NnoM6PbD6/bCVFzd6rKmGbDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"21844de66894801fc125f96abe3873cf123cf05991103513c85fbb605a8c4fda","last_reissued_at":"2026-07-05T08:46:35.582209Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:46:35.582209Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Falcon2-11B Technical Report","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.CL","authors_text":"Ankit Singh, Basma El Amel Boussaha, Giulia Campesan, Hakim Hacid, Hamza Alobeidli, Kirill Fedyanin, Leen Al Qadi, Maksim Velikanov, Mohamed El Amine Seddik, Mohammed Al-Yafeai, Mugariya Farooq, Nilabhra Roy Chowdhury, Quentin Malartic, Reda Alami, Ruxandra Cojocaru, Sanath Narayan, Yasser Abdelaziz Dahou Djilali","submitted_at":"2024-07-20T14:23:15Z","abstract_excerpt":"We introduce Falcon2-11B, a foundation model trained on over five trillion tokens, and its multimodal counterpart, Falcon2-11B-vlm, which is a vision-to-text model. We report our findings during the training of the Falcon2-11B which follows a multi-stage approach where the early stages are distinguished by their context length and a final stage where we use a curated, high-quality dataset. Additionally, we report the effect of doubling the batch size mid-training and how training loss spikes are affected by the learning rate. The downstream performance of the foundation model is evaluated on e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.14885","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/2407.14885/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":"2407.14885","created_at":"2026-07-05T08:46:35.582273+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.14885v1","created_at":"2026-07-05T08:46:35.582273+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.14885","created_at":"2026-07-05T08:46:35.582273+00:00"},{"alias_kind":"pith_short_12","alias_value":"EGCE3ZTISSAB","created_at":"2026-07-05T08:46:35.582273+00:00"},{"alias_kind":"pith_short_16","alias_value":"EGCE3ZTISSAB7QJF","created_at":"2026-07-05T08:46:35.582273+00:00"},{"alias_kind":"pith_short_8","alias_value":"EGCE3ZTI","created_at":"2026-07-05T08:46:35.582273+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.07995","citing_title":"Customer-Agent: Overcoming Context Limitations in Ultra-Long Shopping Trajectories via Tool-Augmented Agents and RLVR","ref_index":83,"is_internal_anchor":false},{"citing_arxiv_id":"2510.24328","citing_title":"Beyond MCQ: An Open-Ended Arabic Cultural QA Benchmark with Dialect Variants","ref_index":15,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EGCE3ZTISSAB7QJF7FVL4ODTZ4","json":"https://pith.science/pith/EGCE3ZTISSAB7QJF7FVL4ODTZ4.json","graph_json":"https://pith.science/api/pith-number/EGCE3ZTISSAB7QJF7FVL4ODTZ4/graph.json","events_json":"https://pith.science/api/pith-number/EGCE3ZTISSAB7QJF7FVL4ODTZ4/events.json","paper":"https://pith.science/paper/EGCE3ZTI"},"agent_actions":{"view_html":"https://pith.science/pith/EGCE3ZTISSAB7QJF7FVL4ODTZ4","download_json":"https://pith.science/pith/EGCE3ZTISSAB7QJF7FVL4ODTZ4.json","view_paper":"https://pith.science/paper/EGCE3ZTI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.14885&json=true","fetch_graph":"https://pith.science/api/pith-number/EGCE3ZTISSAB7QJF7FVL4ODTZ4/graph.json","fetch_events":"https://pith.science/api/pith-number/EGCE3ZTISSAB7QJF7FVL4ODTZ4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EGCE3ZTISSAB7QJF7FVL4ODTZ4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EGCE3ZTISSAB7QJF7FVL4ODTZ4/action/storage_attestation","attest_author":"https://pith.science/pith/EGCE3ZTISSAB7QJF7FVL4ODTZ4/action/author_attestation","sign_citation":"https://pith.science/pith/EGCE3ZTISSAB7QJF7FVL4ODTZ4/action/citation_signature","submit_replication":"https://pith.science/pith/EGCE3ZTISSAB7QJF7FVL4ODTZ4/action/replication_record"}},"created_at":"2026-07-05T08:46:35.582273+00:00","updated_at":"2026-07-05T08:46:35.582273+00:00"}