{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:5YRHPHYHPQ72WIAWHLD42756E7","short_pith_number":"pith:5YRHPHYH","schema_version":"1.0","canonical_sha256":"ee22779f077c3fab20163ac7cd7fbe27d7ebb13b482e62f1ea4565a49c2c9870","source":{"kind":"arxiv","id":"2408.02354","version":3},"attestation_state":"computed","paper":{"title":"RECE: Reduced Cross-Entropy Loss for Large-Catalogue Sequential Recommenders","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.IR","authors_text":"Danil Gusak, Evgeny Frolov, Gleb Mezentsev, Ivan Oseledets","submitted_at":"2024-08-05T10:02:29Z","abstract_excerpt":"Scalability is a major challenge in modern recommender systems. In sequential recommendations, full Cross-Entropy (CE) loss achieves state-of-the-art recommendation quality but consumes excessive GPU memory with large item catalogs, limiting its practicality. Using a GPU-efficient locality-sensitive hashing-like algorithm for approximating large tensor of logits, this paper introduces a novel RECE (REduced Cross-Entropy) loss. RECE significantly reduces memory consumption while allowing one to enjoy the state-of-the-art performance of full CE loss. Experimental results on various datasets show"},"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":"2408.02354","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2024-08-05T10:02:29Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"2585b2a21a686bc2964353608ed9992abeb9cabe68fe68fd2e097666747d749b","abstract_canon_sha256":"8a984eb2954b447fcf9df711ae2ac852d126a06c1c6ffe1530a20f0864068872"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:55:23.664295Z","signature_b64":"S6UGXspWkJeGllB2IyYeUrul4fr/+Ec0oBuS80a8tYz+30HbkLlH/JS5rHzZyyQveEKFKqT4svdxnBEKXX1lCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ee22779f077c3fab20163ac7cd7fbe27d7ebb13b482e62f1ea4565a49c2c9870","last_reissued_at":"2026-07-05T08:55:23.663897Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:55:23.663897Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"RECE: Reduced Cross-Entropy Loss for Large-Catalogue Sequential Recommenders","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.IR","authors_text":"Danil Gusak, Evgeny Frolov, Gleb Mezentsev, Ivan Oseledets","submitted_at":"2024-08-05T10:02:29Z","abstract_excerpt":"Scalability is a major challenge in modern recommender systems. In sequential recommendations, full Cross-Entropy (CE) loss achieves state-of-the-art recommendation quality but consumes excessive GPU memory with large item catalogs, limiting its practicality. Using a GPU-efficient locality-sensitive hashing-like algorithm for approximating large tensor of logits, this paper introduces a novel RECE (REduced Cross-Entropy) loss. RECE significantly reduces memory consumption while allowing one to enjoy the state-of-the-art performance of full CE loss. Experimental results on various datasets show"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.02354","kind":"arxiv","version":3},"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/2408.02354/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":"2408.02354","created_at":"2026-07-05T08:55:23.663951+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.02354v3","created_at":"2026-07-05T08:55:23.663951+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.02354","created_at":"2026-07-05T08:55:23.663951+00:00"},{"alias_kind":"pith_short_12","alias_value":"5YRHPHYHPQ72","created_at":"2026-07-05T08:55:23.663951+00:00"},{"alias_kind":"pith_short_16","alias_value":"5YRHPHYHPQ72WIAW","created_at":"2026-07-05T08:55:23.663951+00:00"},{"alias_kind":"pith_short_8","alias_value":"5YRHPHYH","created_at":"2026-07-05T08:55:23.663951+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/5YRHPHYHPQ72WIAWHLD42756E7","json":"https://pith.science/pith/5YRHPHYHPQ72WIAWHLD42756E7.json","graph_json":"https://pith.science/api/pith-number/5YRHPHYHPQ72WIAWHLD42756E7/graph.json","events_json":"https://pith.science/api/pith-number/5YRHPHYHPQ72WIAWHLD42756E7/events.json","paper":"https://pith.science/paper/5YRHPHYH"},"agent_actions":{"view_html":"https://pith.science/pith/5YRHPHYHPQ72WIAWHLD42756E7","download_json":"https://pith.science/pith/5YRHPHYHPQ72WIAWHLD42756E7.json","view_paper":"https://pith.science/paper/5YRHPHYH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.02354&json=true","fetch_graph":"https://pith.science/api/pith-number/5YRHPHYHPQ72WIAWHLD42756E7/graph.json","fetch_events":"https://pith.science/api/pith-number/5YRHPHYHPQ72WIAWHLD42756E7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5YRHPHYHPQ72WIAWHLD42756E7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5YRHPHYHPQ72WIAWHLD42756E7/action/storage_attestation","attest_author":"https://pith.science/pith/5YRHPHYHPQ72WIAWHLD42756E7/action/author_attestation","sign_citation":"https://pith.science/pith/5YRHPHYHPQ72WIAWHLD42756E7/action/citation_signature","submit_replication":"https://pith.science/pith/5YRHPHYHPQ72WIAWHLD42756E7/action/replication_record"}},"created_at":"2026-07-05T08:55:23.663951+00:00","updated_at":"2026-07-05T08:55:23.663951+00:00"}