{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:EQWLRQYMRUPFS4KJGFZVGM247D","short_pith_number":"pith:EQWLRQYM","schema_version":"1.0","canonical_sha256":"242cb8c30c8d1e597149317353335cf8c7c5156c744806cde9937756293981e5","source":{"kind":"arxiv","id":"2511.06077","version":3},"attestation_state":"computed","paper":{"title":"Make It Long, Keep It Fast: End-to-End 10K Long User Behavior Sequence Modeling for Billion-Scale Douyin Recommendation","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"Stacked cross-attention and shared batching let recommendation models process 10,000-length user histories at production speed.","cross_cats":["cs.IR"],"primary_cat":"cs.LG","authors_text":"Beichuan Zhang, Bo Sun, Feng Zhang, Hangyu Wang, Jia-Qi Yang, Jinan Ni, Lin Guan, Qiwei Chen, Xiaowen Li, Xiao Yang, Xuanyuan Luo, Yi Cheng, Yuhang Qi, Zhifang Fan, Zhishan Zhao","submitted_at":"2025-11-08T17:22:54Z","abstract_excerpt":"Short-video recommenders such as Douyin must exploit extremely long user behavior histories without breaking latency or cost budgets. We present an end-to-end industrial recommender system that scales long-sequence recommendation modeling to 10K-length histories in production. First, we introduce Stacked Target-to-History Cross Attention (STCA), which replaces history self-attention with stacked cross-attention from the target to the history, reducing complexity from quadratic to linear in sequence length and enabling efficient end-to-end training over long user behavior sequences. Second, we "},"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":true},"canonical_record":{"source":{"id":"2511.06077","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-08T17:22:54Z","cross_cats_sorted":["cs.IR"],"title_canon_sha256":"cc635af1a9f45ca110d27edc47511eaba70be6c0f98b8bd6f75e285217ea6697","abstract_canon_sha256":"b4d0db21bf900b36037cdd67e71cb0bbb313470c8197f8e8cced64b2cfe73502"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:05:01.391858Z","signature_b64":"SyVS9GJKTl2XsUIqRzog6+8pRqhmeeia6CvE9csszQ4B6HdPL2kFD+cpPLNXAKZcvY9V3ftgrEmlUIEZhWj8Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"242cb8c30c8d1e597149317353335cf8c7c5156c744806cde9937756293981e5","last_reissued_at":"2026-05-20T01:05:01.391006Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:05:01.391006Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Make It Long, Keep It Fast: End-to-End 10K Long User Behavior Sequence Modeling for Billion-Scale Douyin Recommendation","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"Stacked cross-attention and shared batching let recommendation models process 10,000-length user histories at production speed.","cross_cats":["cs.IR"],"primary_cat":"cs.LG","authors_text":"Beichuan Zhang, Bo Sun, Feng Zhang, Hangyu Wang, Jia-Qi Yang, Jinan Ni, Lin Guan, Qiwei Chen, Xiaowen Li, Xiao Yang, Xuanyuan Luo, Yi Cheng, Yuhang Qi, Zhifang Fan, Zhishan Zhao","submitted_at":"2025-11-08T17:22:54Z","abstract_excerpt":"Short-video recommenders such as Douyin must exploit extremely long user behavior histories without breaking latency or cost budgets. We present an end-to-end industrial recommender system that scales long-sequence recommendation modeling to 10K-length histories in production. First, we introduce Stacked Target-to-History Cross Attention (STCA), which replaces history self-attention with stacked cross-attention from the target to the history, reducing complexity from quadratic to linear in sequence length and enabling efficient end-to-end training over long user behavior sequences. Second, we "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Deployed at full traffic on Douyin, our system delivers significant improvements on key engagement metrics while meeting production latency, demonstrating a practical path to scaling end-to-end long-sequence recommendation to the 10k regime.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the length-extrapolative training strategy (train on shorter windows, infer on 10k) generalizes without performance loss and that stacked target-to-history cross attention captures the necessary preference signals from long histories.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Douyin deploys stacked target-to-history cross attention and request-level batching to scale end-to-end recommendation modeling to 10k-length histories, observing scaling-law gains and live engagement improvements.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Stacked cross-attention and shared batching let recommendation models process 10,000-length user histories at production speed.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"6c9fe21bb1b197fced8f030413e920a1b111de53a1a18b417aeedf10f682fe5d"},"source":{"id":"2511.06077","kind":"arxiv","version":3},"verdict":{"id":"0d78b2d3-d76d-4426-bce8-7c5512172294","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-17T23:25:16.351949Z","strongest_claim":"Deployed at full traffic on Douyin, our system delivers significant improvements on key engagement metrics while meeting production latency, demonstrating a practical path to scaling end-to-end long-sequence recommendation to the 10k regime.","one_line_summary":"Douyin deploys stacked target-to-history cross attention and request-level batching to scale end-to-end recommendation modeling to 10k-length histories, observing scaling-law gains and live engagement improvements.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the length-extrapolative training strategy (train on shorter windows, infer on 10k) generalizes without performance loss and that stacked target-to-history cross attention captures the necessary preference signals from long histories.","pith_extraction_headline":"Stacked cross-attention and shared batching let recommendation models process 10,000-length user histories at production speed."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2511.06077/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":2,"snapshot_sha256":"7a41d85a461f01b63038805dbe2e9b71b94ca760b6a1a14637201f6949c8c6cd"},"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":"2511.06077","created_at":"2026-05-20T01:05:01.391129+00:00"},{"alias_kind":"arxiv_version","alias_value":"2511.06077v3","created_at":"2026-05-20T01:05:01.391129+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.06077","created_at":"2026-05-20T01:05:01.391129+00:00"},{"alias_kind":"pith_short_12","alias_value":"EQWLRQYMRUPF","created_at":"2026-05-20T01:05:01.391129+00:00"},{"alias_kind":"pith_short_16","alias_value":"EQWLRQYMRUPFS4KJ","created_at":"2026-05-20T01:05:01.391129+00:00"},{"alias_kind":"pith_short_8","alias_value":"EQWLRQYM","created_at":"2026-05-20T01:05:01.391129+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":4,"sample":[{"citing_arxiv_id":"2604.23810","citing_title":"Similar Users-Augmented Interest Network","ref_index":20,"is_internal_anchor":true},{"citing_arxiv_id":"2605.04450","citing_title":"One Pool, Two Caches: Adaptive HBM Partitioning for Accelerating Generative Recommender Serving","ref_index":14,"is_internal_anchor":true},{"citing_arxiv_id":"2604.08933","citing_title":"IAT: Instance-As-Token Compression for Historical User Sequence Modeling in Industrial Recommender Systems","ref_index":6,"is_internal_anchor":true},{"citing_arxiv_id":"2605.04450","citing_title":"One Pool, Two Caches: Adaptive HBM Partitioning for Accelerating Generative Recommender Serving","ref_index":15,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EQWLRQYMRUPFS4KJGFZVGM247D","json":"https://pith.science/pith/EQWLRQYMRUPFS4KJGFZVGM247D.json","graph_json":"https://pith.science/api/pith-number/EQWLRQYMRUPFS4KJGFZVGM247D/graph.json","events_json":"https://pith.science/api/pith-number/EQWLRQYMRUPFS4KJGFZVGM247D/events.json","paper":"https://pith.science/paper/EQWLRQYM"},"agent_actions":{"view_html":"https://pith.science/pith/EQWLRQYMRUPFS4KJGFZVGM247D","download_json":"https://pith.science/pith/EQWLRQYMRUPFS4KJGFZVGM247D.json","view_paper":"https://pith.science/paper/EQWLRQYM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2511.06077&json=true","fetch_graph":"https://pith.science/api/pith-number/EQWLRQYMRUPFS4KJGFZVGM247D/graph.json","fetch_events":"https://pith.science/api/pith-number/EQWLRQYMRUPFS4KJGFZVGM247D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EQWLRQYMRUPFS4KJGFZVGM247D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EQWLRQYMRUPFS4KJGFZVGM247D/action/storage_attestation","attest_author":"https://pith.science/pith/EQWLRQYMRUPFS4KJGFZVGM247D/action/author_attestation","sign_citation":"https://pith.science/pith/EQWLRQYMRUPFS4KJGFZVGM247D/action/citation_signature","submit_replication":"https://pith.science/pith/EQWLRQYMRUPFS4KJGFZVGM247D/action/replication_record"}},"created_at":"2026-05-20T01:05:01.391129+00:00","updated_at":"2026-05-20T01:05:01.391129+00:00"}