{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:V6SDHWEEDSRVS74MO6ZLR5PDFW","short_pith_number":"pith:V6SDHWEE","schema_version":"1.0","canonical_sha256":"afa433d8841ca3597f8c77b2b8f5e32d8a7aa82f167399b84337e2dac05dfe7b","source":{"kind":"arxiv","id":"2402.02314","version":3},"attestation_state":"computed","paper":{"title":"Selecting Large Language Model to Fine-tune via Rectified Scaling Law","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Baizhou Huang, Haotian Ye, Haowei Lin, James Zou, Jianzhu Ma, Qinyu Chen, Sujian Li, Xiaojun Wan, Yitao Liang, Zihao Wang","submitted_at":"2024-02-04T01:55:00Z","abstract_excerpt":"The ever-growing ecosystem of LLMs has posed a challenge in selecting the most appropriate pre-trained model to fine-tune amidst a sea of options. Given constrained resources, fine-tuning all models and making selections afterward is unrealistic. In this work, we formulate this resource-constrained selection task into predicting fine-tuning performance and illustrate its natural connection with Scaling Law. Unlike pre-training, we find that the fine-tuning scaling curve includes not just the well-known \"power phase\" but also the previously unobserved \"pre-power phase\". We also explain why exis"},"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":"2402.02314","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-02-04T01:55:00Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"4b199e5265f38f6a21ee9e88a62add5925bbfa06474bf6d43325ec8c01193acb","abstract_canon_sha256":"1b550ce9718eadf0721f09c485e65fb11a125cced00d318d90884750770fe122"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:24:00.706855Z","signature_b64":"AMHTm+u/eLtzxow76Po/eK+L1501k2wlLdk+jGfnhkgpOnL0c1LzT922/uume3kSNhjIeNzwxBl4WF+h6brvBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"afa433d8841ca3597f8c77b2b8f5e32d8a7aa82f167399b84337e2dac05dfe7b","last_reissued_at":"2026-07-05T08:24:00.706346Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:24:00.706346Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Selecting Large Language Model to Fine-tune via Rectified Scaling Law","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Baizhou Huang, Haotian Ye, Haowei Lin, James Zou, Jianzhu Ma, Qinyu Chen, Sujian Li, Xiaojun Wan, Yitao Liang, Zihao Wang","submitted_at":"2024-02-04T01:55:00Z","abstract_excerpt":"The ever-growing ecosystem of LLMs has posed a challenge in selecting the most appropriate pre-trained model to fine-tune amidst a sea of options. Given constrained resources, fine-tuning all models and making selections afterward is unrealistic. In this work, we formulate this resource-constrained selection task into predicting fine-tuning performance and illustrate its natural connection with Scaling Law. Unlike pre-training, we find that the fine-tuning scaling curve includes not just the well-known \"power phase\" but also the previously unobserved \"pre-power phase\". We also explain why exis"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.02314","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/2402.02314/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":"2402.02314","created_at":"2026-07-05T08:24:00.706411+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.02314v3","created_at":"2026-07-05T08:24:00.706411+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.02314","created_at":"2026-07-05T08:24:00.706411+00:00"},{"alias_kind":"pith_short_12","alias_value":"V6SDHWEEDSRV","created_at":"2026-07-05T08:24:00.706411+00:00"},{"alias_kind":"pith_short_16","alias_value":"V6SDHWEEDSRVS74M","created_at":"2026-07-05T08:24:00.706411+00:00"},{"alias_kind":"pith_short_8","alias_value":"V6SDHWEE","created_at":"2026-07-05T08:24:00.706411+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.01727","citing_title":"When Does Generating More Help? Disentangling Fixed-Source Synthesis from Source Expansion in Synthetic Data Scaling","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2412.02125","citing_title":"Preference Goal Tuning: Post-Training as Latent Control for Frozen Policies","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01325","citing_title":"Rethinking Model Selection in VLM Through the Lens of Gromov-Wasserstein Distance","ref_index":27,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/V6SDHWEEDSRVS74MO6ZLR5PDFW","json":"https://pith.science/pith/V6SDHWEEDSRVS74MO6ZLR5PDFW.json","graph_json":"https://pith.science/api/pith-number/V6SDHWEEDSRVS74MO6ZLR5PDFW/graph.json","events_json":"https://pith.science/api/pith-number/V6SDHWEEDSRVS74MO6ZLR5PDFW/events.json","paper":"https://pith.science/paper/V6SDHWEE"},"agent_actions":{"view_html":"https://pith.science/pith/V6SDHWEEDSRVS74MO6ZLR5PDFW","download_json":"https://pith.science/pith/V6SDHWEEDSRVS74MO6ZLR5PDFW.json","view_paper":"https://pith.science/paper/V6SDHWEE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.02314&json=true","fetch_graph":"https://pith.science/api/pith-number/V6SDHWEEDSRVS74MO6ZLR5PDFW/graph.json","fetch_events":"https://pith.science/api/pith-number/V6SDHWEEDSRVS74MO6ZLR5PDFW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V6SDHWEEDSRVS74MO6ZLR5PDFW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V6SDHWEEDSRVS74MO6ZLR5PDFW/action/storage_attestation","attest_author":"https://pith.science/pith/V6SDHWEEDSRVS74MO6ZLR5PDFW/action/author_attestation","sign_citation":"https://pith.science/pith/V6SDHWEEDSRVS74MO6ZLR5PDFW/action/citation_signature","submit_replication":"https://pith.science/pith/V6SDHWEEDSRVS74MO6ZLR5PDFW/action/replication_record"}},"created_at":"2026-07-05T08:24:00.706411+00:00","updated_at":"2026-07-05T08:24:00.706411+00:00"}