{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:WYSNMPY62ZNTYKVJ7JZAFVA7VO","short_pith_number":"pith:WYSNMPY6","schema_version":"1.0","canonical_sha256":"b624d63f1ed65b3c2aa9fa7202d41fabb9308528f554cc21c60f9cf8a21e5246","source":{"kind":"arxiv","id":"2410.17276","version":2},"attestation_state":"computed","paper":{"title":"Evaluating Performance and Bias of Negative Sampling in Large-Scale Sequential Recommendation Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.IR","authors_text":"Arushi Prakash, Dimitrios Bermperidis, Srivas Chennu","submitted_at":"2024-10-08T00:23:17Z","abstract_excerpt":"Large-scale industrial recommendation models predict the most relevant items from catalogs containing millions or billions of options. To train these models efficiently, a small set of irrelevant items (negative samples) is selected from the vast catalog for each relevant item (positive example), helping the model distinguish between relevant and irrelevant items. Choosing the right negative sampling method is a common challenge. We address this by implementing and comparing various negative sampling methods - random, popularity-based, in-batch, mixed, adaptive, and adaptive with mixed variant"},"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":"2410.17276","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2024-10-08T00:23:17Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"2898f35703d01a66d0cfd60572cfdcc51cc036e8012372a8e8018e407a70d2e7","abstract_canon_sha256":"5f48d299e0fb70a8eda7c51a218e1e11aec6113e07ada8ee3840c6c0de9d90f4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:27:25.487138Z","signature_b64":"kth/nvcqTnXPEPAy/lxEgrBjigalrNUS6PfaGZfHcMXQllGn+42HKQ1o3ZuPQRxlzLBF1ck21hY6W9Rhge6yCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b624d63f1ed65b3c2aa9fa7202d41fabb9308528f554cc21c60f9cf8a21e5246","last_reissued_at":"2026-07-05T09:27:25.486608Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:27:25.486608Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evaluating Performance and Bias of Negative Sampling in Large-Scale Sequential Recommendation Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.IR","authors_text":"Arushi Prakash, Dimitrios Bermperidis, Srivas Chennu","submitted_at":"2024-10-08T00:23:17Z","abstract_excerpt":"Large-scale industrial recommendation models predict the most relevant items from catalogs containing millions or billions of options. To train these models efficiently, a small set of irrelevant items (negative samples) is selected from the vast catalog for each relevant item (positive example), helping the model distinguish between relevant and irrelevant items. Choosing the right negative sampling method is a common challenge. We address this by implementing and comparing various negative sampling methods - random, popularity-based, in-batch, mixed, adaptive, and adaptive with mixed variant"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.17276","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/2410.17276/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":"2410.17276","created_at":"2026-07-05T09:27:25.486673+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.17276v2","created_at":"2026-07-05T09:27:25.486673+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.17276","created_at":"2026-07-05T09:27:25.486673+00:00"},{"alias_kind":"pith_short_12","alias_value":"WYSNMPY62ZNT","created_at":"2026-07-05T09:27:25.486673+00:00"},{"alias_kind":"pith_short_16","alias_value":"WYSNMPY62ZNTYKVJ","created_at":"2026-07-05T09:27:25.486673+00:00"},{"alias_kind":"pith_short_8","alias_value":"WYSNMPY6","created_at":"2026-07-05T09:27:25.486673+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2509.09682","citing_title":"Faster and Memory-Efficient Training of Sequential Recommendation Models for Large Catalogs","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17854","citing_title":"Learning over Positive and Negative Edges with Contrastive Message Passing","ref_index":20,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WYSNMPY62ZNTYKVJ7JZAFVA7VO","json":"https://pith.science/pith/WYSNMPY62ZNTYKVJ7JZAFVA7VO.json","graph_json":"https://pith.science/api/pith-number/WYSNMPY62ZNTYKVJ7JZAFVA7VO/graph.json","events_json":"https://pith.science/api/pith-number/WYSNMPY62ZNTYKVJ7JZAFVA7VO/events.json","paper":"https://pith.science/paper/WYSNMPY6"},"agent_actions":{"view_html":"https://pith.science/pith/WYSNMPY62ZNTYKVJ7JZAFVA7VO","download_json":"https://pith.science/pith/WYSNMPY62ZNTYKVJ7JZAFVA7VO.json","view_paper":"https://pith.science/paper/WYSNMPY6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.17276&json=true","fetch_graph":"https://pith.science/api/pith-number/WYSNMPY62ZNTYKVJ7JZAFVA7VO/graph.json","fetch_events":"https://pith.science/api/pith-number/WYSNMPY62ZNTYKVJ7JZAFVA7VO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WYSNMPY62ZNTYKVJ7JZAFVA7VO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WYSNMPY62ZNTYKVJ7JZAFVA7VO/action/storage_attestation","attest_author":"https://pith.science/pith/WYSNMPY62ZNTYKVJ7JZAFVA7VO/action/author_attestation","sign_citation":"https://pith.science/pith/WYSNMPY62ZNTYKVJ7JZAFVA7VO/action/citation_signature","submit_replication":"https://pith.science/pith/WYSNMPY62ZNTYKVJ7JZAFVA7VO/action/replication_record"}},"created_at":"2026-07-05T09:27:25.486673+00:00","updated_at":"2026-07-05T09:27:25.486673+00:00"}