{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:PCEBSND6ARWJBBLGXIVYKQNJ63","short_pith_number":"pith:PCEBSND6","schema_version":"1.0","canonical_sha256":"788819347e046c908566ba2b8541a9f6e606f743e934c1a0d89c5b687fb5e4aa","source":{"kind":"arxiv","id":"2506.02267","version":1},"attestation_state":"computed","paper":{"title":"TransAct V2: Lifelong User Action Sequence Modeling on Pinterest Recommendation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.IR","authors_text":"Dhruvil Deven Badani, Jiajing Xu, Kangnan Li, Kousik Rajesh, Nikil Pancha, Pong Eksombatchai, Saurabh Vishwas Joshi, Xue Xia, Yangyi Lu","submitted_at":"2025-06-02T21:15:20Z","abstract_excerpt":"Modeling user action sequences has become a popular focus in industrial recommendation system research, particularly for Click-Through Rate (CTR) prediction tasks. However, industry-scale CTR models often rely on short user sequences, limiting their ability to capture long-term behavior. Additionally, these models typically lack an integrated action-prediction task within a point-wise ranking framework, reducing their predictive power. They also rarely address the infrastructure challenges involved in efficiently serving large-scale sequential models. In this paper, we introduce TransAct V2, a"},"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":"2506.02267","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2025-06-02T21:15:20Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8f1af2d2ce1bc635ca58248a480d027faccddb208e577ec730676eb8749cb019","abstract_canon_sha256":"849fbf64a2583196a257fc0f1476b4b25b7b800667bb99f4e919e29a31b7a76b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:14:35.589365Z","signature_b64":"xueZcBC9UYUNM4jeAKGB7bCRVijmcIdaBBkoID6Zc1AX3vqG/zgalhp8v0eSd83KShe0P8P9aDiIER50gUVvDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"788819347e046c908566ba2b8541a9f6e606f743e934c1a0d89c5b687fb5e4aa","last_reissued_at":"2026-07-05T11:14:35.588822Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:14:35.588822Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TransAct V2: Lifelong User Action Sequence Modeling on Pinterest Recommendation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.IR","authors_text":"Dhruvil Deven Badani, Jiajing Xu, Kangnan Li, Kousik Rajesh, Nikil Pancha, Pong Eksombatchai, Saurabh Vishwas Joshi, Xue Xia, Yangyi Lu","submitted_at":"2025-06-02T21:15:20Z","abstract_excerpt":"Modeling user action sequences has become a popular focus in industrial recommendation system research, particularly for Click-Through Rate (CTR) prediction tasks. However, industry-scale CTR models often rely on short user sequences, limiting their ability to capture long-term behavior. Additionally, these models typically lack an integrated action-prediction task within a point-wise ranking framework, reducing their predictive power. They also rarely address the infrastructure challenges involved in efficiently serving large-scale sequential models. In this paper, we introduce TransAct V2, a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.02267","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/2506.02267/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":"2506.02267","created_at":"2026-07-05T11:14:35.588906+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.02267v1","created_at":"2026-07-05T11:14:35.588906+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.02267","created_at":"2026-07-05T11:14:35.588906+00:00"},{"alias_kind":"pith_short_12","alias_value":"PCEBSND6ARWJ","created_at":"2026-07-05T11:14:35.588906+00:00"},{"alias_kind":"pith_short_16","alias_value":"PCEBSND6ARWJBBLG","created_at":"2026-07-05T11:14:35.588906+00:00"},{"alias_kind":"pith_short_8","alias_value":"PCEBSND6","created_at":"2026-07-05T11:14:35.588906+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2511.06077","citing_title":"Make It Long, Keep It Fast: End-to-End 10K Long User Behavior Sequence Modeling for Billion-Scale Douyin Recommendation","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2511.06077","citing_title":"Make It Long, Keep It Fast: End-to-End 10K Long User Behavior Sequence Modeling for Billion-Scale Douyin Recommendation","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08649","citing_title":"PRAGMA: Revolut Foundation Model","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PCEBSND6ARWJBBLGXIVYKQNJ63","json":"https://pith.science/pith/PCEBSND6ARWJBBLGXIVYKQNJ63.json","graph_json":"https://pith.science/api/pith-number/PCEBSND6ARWJBBLGXIVYKQNJ63/graph.json","events_json":"https://pith.science/api/pith-number/PCEBSND6ARWJBBLGXIVYKQNJ63/events.json","paper":"https://pith.science/paper/PCEBSND6"},"agent_actions":{"view_html":"https://pith.science/pith/PCEBSND6ARWJBBLGXIVYKQNJ63","download_json":"https://pith.science/pith/PCEBSND6ARWJBBLGXIVYKQNJ63.json","view_paper":"https://pith.science/paper/PCEBSND6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.02267&json=true","fetch_graph":"https://pith.science/api/pith-number/PCEBSND6ARWJBBLGXIVYKQNJ63/graph.json","fetch_events":"https://pith.science/api/pith-number/PCEBSND6ARWJBBLGXIVYKQNJ63/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PCEBSND6ARWJBBLGXIVYKQNJ63/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PCEBSND6ARWJBBLGXIVYKQNJ63/action/storage_attestation","attest_author":"https://pith.science/pith/PCEBSND6ARWJBBLGXIVYKQNJ63/action/author_attestation","sign_citation":"https://pith.science/pith/PCEBSND6ARWJBBLGXIVYKQNJ63/action/citation_signature","submit_replication":"https://pith.science/pith/PCEBSND6ARWJBBLGXIVYKQNJ63/action/replication_record"}},"created_at":"2026-07-05T11:14:35.588906+00:00","updated_at":"2026-07-05T11:14:35.588906+00:00"}