{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7GRFZS5G3NOSD42IIHAY7BAXNG","short_pith_number":"pith:7GRFZS5G","schema_version":"1.0","canonical_sha256":"f9a25ccba6db5d21f34841c18f841769ad9a1fed9742cb831c80c607f062bea0","source":{"kind":"arxiv","id":"2607.24015","version":1},"attestation_state":"computed","paper":{"title":"Mosaic: A Fleet of User Embedding Specialists for Recommendation at Meta","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Aakarsha Agarwal, Christina You, Hong Li, Hrishikesh Paranjape, John Zhiyuan Zheng, Michael Jiayuan He, Xiangyang Mou, Xian Sun, Yujunrong Ma","submitted_at":"2026-07-27T05:25:50Z","abstract_excerpt":"User representation is one of the highest-leverage modeling problems in industrial recommendation systems: a single advancement in how users are encoded can propagate across retrieval, ranking, and integrity tasks at platform scale. Prior industrial user representation work builds either a single user model that emits one or more embedding vectors or a shared backbone with task-specific adaptation. In this paper, we present Mosaic, a foundational user modeling platform that employs a fleet of specialists to learn user embeddings. The fleet comprises four architecturally diverse model families "},"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":"2607.24015","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-07-27T05:25:50Z","cross_cats_sorted":[],"title_canon_sha256":"be2b367ca87ea16f8ba73531670155842204509d1f2822eab404590f8dcb5b9e","abstract_canon_sha256":"cc3364b621e5b0b0448c44c50859e80bcf7d5a47238446e3610b35bd8c726ffe"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-28T01:23:41.209939Z","signature_b64":"bCzKUr7dhwNFOoQD2jEaQUUzQ3o3k9yojNUVeasIPUzzf2b1nqKaHxP0BcVDoUJgmM1v7TVGbjnOCDcUk5TuCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f9a25ccba6db5d21f34841c18f841769ad9a1fed9742cb831c80c607f062bea0","last_reissued_at":"2026-07-28T01:23:41.209110Z","signature_status":"signed_v1","first_computed_at":"2026-07-28T01:23:41.209110Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mosaic: A Fleet of User Embedding Specialists for Recommendation at Meta","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Aakarsha Agarwal, Christina You, Hong Li, Hrishikesh Paranjape, John Zhiyuan Zheng, Michael Jiayuan He, Xiangyang Mou, Xian Sun, Yujunrong Ma","submitted_at":"2026-07-27T05:25:50Z","abstract_excerpt":"User representation is one of the highest-leverage modeling problems in industrial recommendation systems: a single advancement in how users are encoded can propagate across retrieval, ranking, and integrity tasks at platform scale. Prior industrial user representation work builds either a single user model that emits one or more embedding vectors or a shared backbone with task-specific adaptation. In this paper, we present Mosaic, a foundational user modeling platform that employs a fleet of specialists to learn user embeddings. The fleet comprises four architecturally diverse model families "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.24015","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/2607.24015/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":"2607.24015","created_at":"2026-07-28T01:23:41.209543+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.24015v1","created_at":"2026-07-28T01:23:41.209543+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.24015","created_at":"2026-07-28T01:23:41.209543+00:00"},{"alias_kind":"pith_short_12","alias_value":"7GRFZS5G3NOS","created_at":"2026-07-28T01:23:41.209543+00:00"},{"alias_kind":"pith_short_16","alias_value":"7GRFZS5G3NOSD42I","created_at":"2026-07-28T01:23:41.209543+00:00"},{"alias_kind":"pith_short_8","alias_value":"7GRFZS5G","created_at":"2026-07-28T01:23:41.209543+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/7GRFZS5G3NOSD42IIHAY7BAXNG","json":"https://pith.science/pith/7GRFZS5G3NOSD42IIHAY7BAXNG.json","graph_json":"https://pith.science/api/pith-number/7GRFZS5G3NOSD42IIHAY7BAXNG/graph.json","events_json":"https://pith.science/api/pith-number/7GRFZS5G3NOSD42IIHAY7BAXNG/events.json","paper":"https://pith.science/paper/7GRFZS5G"},"agent_actions":{"view_html":"https://pith.science/pith/7GRFZS5G3NOSD42IIHAY7BAXNG","download_json":"https://pith.science/pith/7GRFZS5G3NOSD42IIHAY7BAXNG.json","view_paper":"https://pith.science/paper/7GRFZS5G","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.24015&json=true","fetch_graph":"https://pith.science/api/pith-number/7GRFZS5G3NOSD42IIHAY7BAXNG/graph.json","fetch_events":"https://pith.science/api/pith-number/7GRFZS5G3NOSD42IIHAY7BAXNG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7GRFZS5G3NOSD42IIHAY7BAXNG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7GRFZS5G3NOSD42IIHAY7BAXNG/action/storage_attestation","attest_author":"https://pith.science/pith/7GRFZS5G3NOSD42IIHAY7BAXNG/action/author_attestation","sign_citation":"https://pith.science/pith/7GRFZS5G3NOSD42IIHAY7BAXNG/action/citation_signature","submit_replication":"https://pith.science/pith/7GRFZS5G3NOSD42IIHAY7BAXNG/action/replication_record"}},"created_at":"2026-07-28T01:23:41.209543+00:00","updated_at":"2026-07-28T01:23:41.209543+00:00"}