{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:VVIQCSUPVAPTBTB4YBP3W573WH","short_pith_number":"pith:VVIQCSUP","schema_version":"1.0","canonical_sha256":"ad51014a8fa81f30cc3cc05fbb77fbb1c674b58392871ea204b5f0270fe11597","source":{"kind":"arxiv","id":"2605.23191","version":1},"attestation_state":"computed","paper":{"title":"Expand More, Shrink Less: Shaping Effective-Rank Dynamics for Dense Scaling in Recommendation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.IR","cs.NA","math.NA"],"primary_cat":"cs.LG","authors_text":"Chao Zhou, Gengsheng Xue, Guoming Li, Haijie Gu, Jin Chen, Junwei Pan, Menglin Yang, Shangyu Zhang, Shudong Huang, Wentao Ning","submitted_at":"2026-05-22T03:17:29Z","abstract_excerpt":"Scaling recommendation models is a central challenge in recommender systems. Recently, RankMixer has emerged as an effective solution, operating on a unified token representation and alternating between token mixing and per-token feedforward networks (P-FFNs) to achieve scalable performance. However, RankMixer suffers from \\textit{embedding collapse}, where learned representations have low effective rank, limiting expressivity and underutilizing the expanded representation space. Through empirical analysis and theoretical insights, we identify rigid token mixing and P-FFN modules as the primar"},"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":"2605.23191","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-22T03:17:29Z","cross_cats_sorted":["cs.IR","cs.NA","math.NA"],"title_canon_sha256":"7278d30625b74511c2fd27c2c7a4cc0ec06ba57d57d635a5013fd4d5db56f661","abstract_canon_sha256":"8f31f56dc228df2bfabebabf1ce7912a4314f2ee631d57480fa9429d229fe1a9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-25T02:01:42.384155Z","signature_b64":"lDPGS1BVWr501TQVMjFYH3mZlPRUsW4qQg0RhTvg36dtfnz4uLZUM8FSzMrjI30bIhS8JJ5gzWLP6wyaRj7IDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ad51014a8fa81f30cc3cc05fbb77fbb1c674b58392871ea204b5f0270fe11597","last_reissued_at":"2026-05-25T02:01:42.383583Z","signature_status":"signed_v1","first_computed_at":"2026-05-25T02:01:42.383583Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Expand More, Shrink Less: Shaping Effective-Rank Dynamics for Dense Scaling in Recommendation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.IR","cs.NA","math.NA"],"primary_cat":"cs.LG","authors_text":"Chao Zhou, Gengsheng Xue, Guoming Li, Haijie Gu, Jin Chen, Junwei Pan, Menglin Yang, Shangyu Zhang, Shudong Huang, Wentao Ning","submitted_at":"2026-05-22T03:17:29Z","abstract_excerpt":"Scaling recommendation models is a central challenge in recommender systems. Recently, RankMixer has emerged as an effective solution, operating on a unified token representation and alternating between token mixing and per-token feedforward networks (P-FFNs) to achieve scalable performance. However, RankMixer suffers from \\textit{embedding collapse}, where learned representations have low effective rank, limiting expressivity and underutilizing the expanded representation space. Through empirical analysis and theoretical insights, we identify rigid token mixing and P-FFN modules as the primar"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.23191","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/2605.23191/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":"2605.23191","created_at":"2026-05-25T02:01:42.383699+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.23191v1","created_at":"2026-05-25T02:01:42.383699+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.23191","created_at":"2026-05-25T02:01:42.383699+00:00"},{"alias_kind":"pith_short_12","alias_value":"VVIQCSUPVAPT","created_at":"2026-05-25T02:01:42.383699+00:00"},{"alias_kind":"pith_short_16","alias_value":"VVIQCSUPVAPTBTB4","created_at":"2026-05-25T02:01:42.383699+00:00"},{"alias_kind":"pith_short_8","alias_value":"VVIQCSUP","created_at":"2026-05-25T02:01:42.383699+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/VVIQCSUPVAPTBTB4YBP3W573WH","json":"https://pith.science/pith/VVIQCSUPVAPTBTB4YBP3W573WH.json","graph_json":"https://pith.science/api/pith-number/VVIQCSUPVAPTBTB4YBP3W573WH/graph.json","events_json":"https://pith.science/api/pith-number/VVIQCSUPVAPTBTB4YBP3W573WH/events.json","paper":"https://pith.science/paper/VVIQCSUP"},"agent_actions":{"view_html":"https://pith.science/pith/VVIQCSUPVAPTBTB4YBP3W573WH","download_json":"https://pith.science/pith/VVIQCSUPVAPTBTB4YBP3W573WH.json","view_paper":"https://pith.science/paper/VVIQCSUP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.23191&json=true","fetch_graph":"https://pith.science/api/pith-number/VVIQCSUPVAPTBTB4YBP3W573WH/graph.json","fetch_events":"https://pith.science/api/pith-number/VVIQCSUPVAPTBTB4YBP3W573WH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VVIQCSUPVAPTBTB4YBP3W573WH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VVIQCSUPVAPTBTB4YBP3W573WH/action/storage_attestation","attest_author":"https://pith.science/pith/VVIQCSUPVAPTBTB4YBP3W573WH/action/author_attestation","sign_citation":"https://pith.science/pith/VVIQCSUPVAPTBTB4YBP3W573WH/action/citation_signature","submit_replication":"https://pith.science/pith/VVIQCSUPVAPTBTB4YBP3W573WH/action/replication_record"}},"created_at":"2026-05-25T02:01:42.383699+00:00","updated_at":"2026-05-25T02:01:42.383699+00:00"}