{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:QOWDQJIFWIB53AZSLPZCJNB7WO","short_pith_number":"pith:QOWDQJIF","schema_version":"1.0","canonical_sha256":"83ac382505b203dd83325bf224b43fb3b37500646bea8949e90a27527d4b0797","source":{"kind":"arxiv","id":"2507.10613","version":1},"attestation_state":"computed","paper":{"title":"Sub-Scaling Laws: On the Role of Data Density and Training Strategies in LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jingang Wang, Junxian He, Shiqi Chen, Siqi Wang, Teng Xiao, Xunliang Cai, Yudong Wang, Zhengyu Chen","submitted_at":"2025-07-13T15:15:24Z","abstract_excerpt":"Traditional scaling laws in natural language processing suggest that increasing model size and training data enhances performance. However, recent studies reveal deviations, particularly in large language models, where performance improvements decelerate, which is a phenomenon known as sub-scaling. This paper revisits these scaling laws by examining the impact of data quality and training strategies on model performance. Through extensive empirical analysis of over 400 models, we identify high data density and non-optimal resource allocation as key factors contributing to sub-scaling. High dat"},"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":"2507.10613","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-07-13T15:15:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"bd3bdfe60989ec1e5691a4d02bb5785492c0f2808a2729f0d459eac00e473926","abstract_canon_sha256":"70fdb218b8266b97b2a7270f7a0f9ada615467d664e964ce46330ff3d5fb962d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:36:57.606386Z","signature_b64":"rr8+yjNo0koTW/hAgTs/wQ26WMD8g2CJKvsZLJ7Btdf53d7lOKbNUKreJFAERk3NXzHQe2xNJK5ZH09BS92/Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"83ac382505b203dd83325bf224b43fb3b37500646bea8949e90a27527d4b0797","last_reissued_at":"2026-07-05T11:36:57.605811Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:36:57.605811Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Sub-Scaling Laws: On the Role of Data Density and Training Strategies in LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jingang Wang, Junxian He, Shiqi Chen, Siqi Wang, Teng Xiao, Xunliang Cai, Yudong Wang, Zhengyu Chen","submitted_at":"2025-07-13T15:15:24Z","abstract_excerpt":"Traditional scaling laws in natural language processing suggest that increasing model size and training data enhances performance. However, recent studies reveal deviations, particularly in large language models, where performance improvements decelerate, which is a phenomenon known as sub-scaling. This paper revisits these scaling laws by examining the impact of data quality and training strategies on model performance. Through extensive empirical analysis of over 400 models, we identify high data density and non-optimal resource allocation as key factors contributing to sub-scaling. High dat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.10613","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/2507.10613/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":"2507.10613","created_at":"2026-07-05T11:36:57.605871+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.10613v1","created_at":"2026-07-05T11:36:57.605871+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.10613","created_at":"2026-07-05T11:36:57.605871+00:00"},{"alias_kind":"pith_short_12","alias_value":"QOWDQJIFWIB5","created_at":"2026-07-05T11:36:57.605871+00:00"},{"alias_kind":"pith_short_16","alias_value":"QOWDQJIFWIB53AZS","created_at":"2026-07-05T11:36:57.605871+00:00"},{"alias_kind":"pith_short_8","alias_value":"QOWDQJIF","created_at":"2026-07-05T11:36:57.605871+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/QOWDQJIFWIB53AZSLPZCJNB7WO","json":"https://pith.science/pith/QOWDQJIFWIB53AZSLPZCJNB7WO.json","graph_json":"https://pith.science/api/pith-number/QOWDQJIFWIB53AZSLPZCJNB7WO/graph.json","events_json":"https://pith.science/api/pith-number/QOWDQJIFWIB53AZSLPZCJNB7WO/events.json","paper":"https://pith.science/paper/QOWDQJIF"},"agent_actions":{"view_html":"https://pith.science/pith/QOWDQJIFWIB53AZSLPZCJNB7WO","download_json":"https://pith.science/pith/QOWDQJIFWIB53AZSLPZCJNB7WO.json","view_paper":"https://pith.science/paper/QOWDQJIF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.10613&json=true","fetch_graph":"https://pith.science/api/pith-number/QOWDQJIFWIB53AZSLPZCJNB7WO/graph.json","fetch_events":"https://pith.science/api/pith-number/QOWDQJIFWIB53AZSLPZCJNB7WO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QOWDQJIFWIB53AZSLPZCJNB7WO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QOWDQJIFWIB53AZSLPZCJNB7WO/action/storage_attestation","attest_author":"https://pith.science/pith/QOWDQJIFWIB53AZSLPZCJNB7WO/action/author_attestation","sign_citation":"https://pith.science/pith/QOWDQJIFWIB53AZSLPZCJNB7WO/action/citation_signature","submit_replication":"https://pith.science/pith/QOWDQJIFWIB53AZSLPZCJNB7WO/action/replication_record"}},"created_at":"2026-07-05T11:36:57.605871+00:00","updated_at":"2026-07-05T11:36:57.605871+00:00"}