{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:3KHDFAASKYXDJVZ7QBIR7TARLW","short_pith_number":"pith:3KHDFAAS","schema_version":"1.0","canonical_sha256":"da8e328012562e34d73f80511fcc115dbff6ad185a51072aef230df5a291569d","source":{"kind":"arxiv","id":"2508.15486","version":2},"attestation_state":"computed","paper":{"title":"LongRetriever: Towards Ultra-Long Sequence based Candidate Retrieval for Recommendation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Di Wu, Qin Ren, Xijun Xiao, Yuchao Zheng, Zheng Chai","submitted_at":"2025-08-21T12:06:36Z","abstract_excerpt":"Precisely modeling user ultra-long sequences is critical for industrial recommender systems. Current approaches predominantly focus on leveraging ultra-long sequences in the ranking stage, whereas research for the candidate retrieval stage remains under-explored. This paper presents LongRetriever, a practical framework for incorporating ultra-long sequences into the retrieval stage of recommenders. Specifically, we propose in-context training and multi-context retrieval, which enable candidate-specific interaction between user sequence and candidate item, and ensure training-serving consistenc"},"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":"2508.15486","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2025-08-21T12:06:36Z","cross_cats_sorted":[],"title_canon_sha256":"c41a3c57dcb74e66e4420218bf8d8b14510f94b75c9baf0bd5240a752158c432","abstract_canon_sha256":"7ee05bca77e0323834e0a09f20eed61b0bd136d2e2646af750131d7c2cdc3083"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:58:36.190038Z","signature_b64":"S20Xui37TNd+VmdCRoyv+duFUTM632aNZTkOFhixxS6fHcUPFAgTfjDypykJhORbEJktQkEq3ZnPARyBfZjuDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"da8e328012562e34d73f80511fcc115dbff6ad185a51072aef230df5a291569d","last_reissued_at":"2026-07-05T11:58:36.189562Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:58:36.189562Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LongRetriever: Towards Ultra-Long Sequence based Candidate Retrieval for Recommendation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Di Wu, Qin Ren, Xijun Xiao, Yuchao Zheng, Zheng Chai","submitted_at":"2025-08-21T12:06:36Z","abstract_excerpt":"Precisely modeling user ultra-long sequences is critical for industrial recommender systems. Current approaches predominantly focus on leveraging ultra-long sequences in the ranking stage, whereas research for the candidate retrieval stage remains under-explored. This paper presents LongRetriever, a practical framework for incorporating ultra-long sequences into the retrieval stage of recommenders. Specifically, we propose in-context training and multi-context retrieval, which enable candidate-specific interaction between user sequence and candidate item, and ensure training-serving consistenc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.15486","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/2508.15486/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":"2508.15486","created_at":"2026-07-05T11:58:36.189621+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.15486v2","created_at":"2026-07-05T11:58:36.189621+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.15486","created_at":"2026-07-05T11:58:36.189621+00:00"},{"alias_kind":"pith_short_12","alias_value":"3KHDFAASKYXD","created_at":"2026-07-05T11:58:36.189621+00:00"},{"alias_kind":"pith_short_16","alias_value":"3KHDFAASKYXDJVZ7","created_at":"2026-07-05T11:58:36.189621+00:00"},{"alias_kind":"pith_short_8","alias_value":"3KHDFAAS","created_at":"2026-07-05T11:58:36.189621+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.29970","citing_title":"From Extraction to Navigation: Progressive Retrieval with Indirectly Infinite Depth","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24806","citing_title":"Versioned Late Materialization for Ultra-Long Sequence Training in Recommendation Systems at Scale","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3KHDFAASKYXDJVZ7QBIR7TARLW","json":"https://pith.science/pith/3KHDFAASKYXDJVZ7QBIR7TARLW.json","graph_json":"https://pith.science/api/pith-number/3KHDFAASKYXDJVZ7QBIR7TARLW/graph.json","events_json":"https://pith.science/api/pith-number/3KHDFAASKYXDJVZ7QBIR7TARLW/events.json","paper":"https://pith.science/paper/3KHDFAAS"},"agent_actions":{"view_html":"https://pith.science/pith/3KHDFAASKYXDJVZ7QBIR7TARLW","download_json":"https://pith.science/pith/3KHDFAASKYXDJVZ7QBIR7TARLW.json","view_paper":"https://pith.science/paper/3KHDFAAS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.15486&json=true","fetch_graph":"https://pith.science/api/pith-number/3KHDFAASKYXDJVZ7QBIR7TARLW/graph.json","fetch_events":"https://pith.science/api/pith-number/3KHDFAASKYXDJVZ7QBIR7TARLW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3KHDFAASKYXDJVZ7QBIR7TARLW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3KHDFAASKYXDJVZ7QBIR7TARLW/action/storage_attestation","attest_author":"https://pith.science/pith/3KHDFAASKYXDJVZ7QBIR7TARLW/action/author_attestation","sign_citation":"https://pith.science/pith/3KHDFAASKYXDJVZ7QBIR7TARLW/action/citation_signature","submit_replication":"https://pith.science/pith/3KHDFAASKYXDJVZ7QBIR7TARLW/action/replication_record"}},"created_at":"2026-07-05T11:58:36.189621+00:00","updated_at":"2026-07-05T11:58:36.189621+00:00"}