{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:YEM6NZCP54YQZU4LQRGVX4JVHA","short_pith_number":"pith:YEM6NZCP","schema_version":"1.0","canonical_sha256":"c119e6e44fef310cd38b844d5bf135382d0906d42163226ee5e33b26289f7cc8","source":{"kind":"arxiv","id":"2506.10972","version":3},"attestation_state":"computed","paper":{"title":"Predictable Scale: Part II, Farseer: A Refined Scaling Law in Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Daxin Jiang, Haoying Wang, Houyi Li, Ning Ding, Qiufeng Wang, Shijie Xuyang, Shuigeng Zhou, Wenzhen Zheng, Xiangyu Zhang, Zhenyu Ding, Zili Wang","submitted_at":"2025-06-12T17:59:23Z","abstract_excerpt":"Training Large Language Models (LLMs) is prohibitively expensive, creating a critical scaling gap where insights from small-scale experiments often fail to transfer to resource-intensive production systems, thereby hindering efficient innovation. To bridge this, we introduce Farseer, a novel and refined scaling law offering enhanced predictive accuracy across scales. By systematically constructing a model loss surface $L(N,D)$, Farseer achieves a significantly better fit to empirical data than prior laws (e.g., Chinchilla's law). Our methodology yields accurate, robust, and highly generalizabl"},"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":"2506.10972","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-06-12T17:59:23Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"fe1cab30521e7350998cd66760423e2c04bff6399e1544726f24cb46a7111872","abstract_canon_sha256":"b7b16602b5c0a017972db8d9c129b912bad5a6f2322f0e0921855dfa4275285d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:38:07.643489Z","signature_b64":"TWlh9+7NoeV4S5NSdtgxyXo/9qH3b9EW6zcDStY2EMJIRvxvd0ptnhJW6yBO01H1+UkqJHK4274fnO76kgY7BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c119e6e44fef310cd38b844d5bf135382d0906d42163226ee5e33b26289f7cc8","last_reissued_at":"2026-07-05T11:38:07.642824Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:38:07.642824Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Predictable Scale: Part II, Farseer: A Refined Scaling Law in Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Daxin Jiang, Haoying Wang, Houyi Li, Ning Ding, Qiufeng Wang, Shijie Xuyang, Shuigeng Zhou, Wenzhen Zheng, Xiangyu Zhang, Zhenyu Ding, Zili Wang","submitted_at":"2025-06-12T17:59:23Z","abstract_excerpt":"Training Large Language Models (LLMs) is prohibitively expensive, creating a critical scaling gap where insights from small-scale experiments often fail to transfer to resource-intensive production systems, thereby hindering efficient innovation. To bridge this, we introduce Farseer, a novel and refined scaling law offering enhanced predictive accuracy across scales. By systematically constructing a model loss surface $L(N,D)$, Farseer achieves a significantly better fit to empirical data than prior laws (e.g., Chinchilla's law). Our methodology yields accurate, robust, and highly generalizabl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.10972","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/2506.10972/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":"2506.10972","created_at":"2026-07-05T11:38:07.642894+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.10972v3","created_at":"2026-07-05T11:38:07.642894+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.10972","created_at":"2026-07-05T11:38:07.642894+00:00"},{"alias_kind":"pith_short_12","alias_value":"YEM6NZCP54YQ","created_at":"2026-07-05T11:38:07.642894+00:00"},{"alias_kind":"pith_short_16","alias_value":"YEM6NZCP54YQZU4L","created_at":"2026-07-05T11:38:07.642894+00:00"},{"alias_kind":"pith_short_8","alias_value":"YEM6NZCP","created_at":"2026-07-05T11:38:07.642894+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25008","citing_title":"Neural Scaling Universality: If Exponents Are Fixed, Time to Understand Coefficients","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02794","citing_title":"Scaling Laws for Neural-Network Quantum States","ref_index":49,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YEM6NZCP54YQZU4LQRGVX4JVHA","json":"https://pith.science/pith/YEM6NZCP54YQZU4LQRGVX4JVHA.json","graph_json":"https://pith.science/api/pith-number/YEM6NZCP54YQZU4LQRGVX4JVHA/graph.json","events_json":"https://pith.science/api/pith-number/YEM6NZCP54YQZU4LQRGVX4JVHA/events.json","paper":"https://pith.science/paper/YEM6NZCP"},"agent_actions":{"view_html":"https://pith.science/pith/YEM6NZCP54YQZU4LQRGVX4JVHA","download_json":"https://pith.science/pith/YEM6NZCP54YQZU4LQRGVX4JVHA.json","view_paper":"https://pith.science/paper/YEM6NZCP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.10972&json=true","fetch_graph":"https://pith.science/api/pith-number/YEM6NZCP54YQZU4LQRGVX4JVHA/graph.json","fetch_events":"https://pith.science/api/pith-number/YEM6NZCP54YQZU4LQRGVX4JVHA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YEM6NZCP54YQZU4LQRGVX4JVHA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YEM6NZCP54YQZU4LQRGVX4JVHA/action/storage_attestation","attest_author":"https://pith.science/pith/YEM6NZCP54YQZU4LQRGVX4JVHA/action/author_attestation","sign_citation":"https://pith.science/pith/YEM6NZCP54YQZU4LQRGVX4JVHA/action/citation_signature","submit_replication":"https://pith.science/pith/YEM6NZCP54YQZU4LQRGVX4JVHA/action/replication_record"}},"created_at":"2026-07-05T11:38:07.642894+00:00","updated_at":"2026-07-05T11:38:07.642894+00:00"}