{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:EHVUIWH5KI7W26ODTL3DG5DFCK","short_pith_number":"pith:EHVUIWH5","schema_version":"1.0","canonical_sha256":"21eb4458fd523f6d79c39af633746512907c2785968c5d749f5cf5873350cc5a","source":{"kind":"arxiv","id":"2102.01293","version":1},"attestation_state":"computed","paper":{"title":"Scaling Laws for Transfer","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Danny Hernandez, Jared Kaplan, Sam McCandlish, Tom Henighan","submitted_at":"2021-02-02T04:07:38Z","abstract_excerpt":"We study empirical scaling laws for transfer learning between distributions in an unsupervised, fine-tuning setting. When we train increasingly large neural networks from-scratch on a fixed-size dataset, they eventually become data-limited and stop improving in performance (cross-entropy loss). When we do the same for models pre-trained on a large language dataset, the slope in performance gains is merely reduced rather than going to zero. We calculate the effective data \"transferred\" from pre-training by determining how much data a transformer of the same size would have required to achieve t"},"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":"2102.01293","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-02-02T04:07:38Z","cross_cats_sorted":[],"title_canon_sha256":"93f5bcc2748fe85792d3a21f550055ab92853970fd2b2b75b653ac4e91188453","abstract_canon_sha256":"691a3905a3d49cc7220e73b78a7d10282a2113bb254d3ec4e391d5293b2e40f4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:52:06.251329Z","signature_b64":"DUJ/fA3usiohgR0iweVEoBcmbihyLsiYVBEF+5e4QeKFvcOMgw9jtk38c0MFfkCCRcrWcHsvP+CpymSX5l9UBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"21eb4458fd523f6d79c39af633746512907c2785968c5d749f5cf5873350cc5a","last_reissued_at":"2026-05-18T00:52:06.250715Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:52:06.250715Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Scaling Laws for Transfer","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Danny Hernandez, Jared Kaplan, Sam McCandlish, Tom Henighan","submitted_at":"2021-02-02T04:07:38Z","abstract_excerpt":"We study empirical scaling laws for transfer learning between distributions in an unsupervised, fine-tuning setting. When we train increasingly large neural networks from-scratch on a fixed-size dataset, they eventually become data-limited and stop improving in performance (cross-entropy loss). When we do the same for models pre-trained on a large language dataset, the slope in performance gains is merely reduced rather than going to zero. We calculate the effective data \"transferred\" from pre-training by determining how much data a transformer of the same size would have required to achieve t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.01293","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":""},"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":"2102.01293","created_at":"2026-05-18T00:52:06.250826+00:00"},{"alias_kind":"arxiv_version","alias_value":"2102.01293v1","created_at":"2026-05-18T00:52:06.250826+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.01293","created_at":"2026-05-18T00:52:06.250826+00:00"},{"alias_kind":"pith_short_12","alias_value":"EHVUIWH5KI7W","created_at":"2026-05-18T12:33:33.725879+00:00"},{"alias_kind":"pith_short_16","alias_value":"EHVUIWH5KI7W26OD","created_at":"2026-05-18T12:33:33.725879+00:00"},{"alias_kind":"pith_short_8","alias_value":"EHVUIWH5","created_at":"2026-05-18T12:33:33.725879+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":19,"internal_anchor_count":19,"sample":[{"citing_arxiv_id":"2303.09540","citing_title":"SemDeDup: Data-efficient learning at web-scale through semantic deduplication","ref_index":6,"is_internal_anchor":true},{"citing_arxiv_id":"2305.16264","citing_title":"Scaling Data-Constrained Language Models","ref_index":44,"is_internal_anchor":true},{"citing_arxiv_id":"2207.14255","citing_title":"Efficient Training of Language Models to Fill in the Middle","ref_index":113,"is_internal_anchor":true},{"citing_arxiv_id":"2205.10487","citing_title":"Scaling Laws and Interpretability of Learning from Repeated Data","ref_index":56,"is_internal_anchor":true},{"citing_arxiv_id":"2405.14782","citing_title":"Lessons from the Trenches on Reproducible Evaluation of Language Models","ref_index":27,"is_internal_anchor":true},{"citing_arxiv_id":"2506.02153","citing_title":"Small Language Models are the Future of Agentic AI","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"2304.01373","citing_title":"Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling","ref_index":120,"is_internal_anchor":true},{"citing_arxiv_id":"2303.17564","citing_title":"BloombergGPT: A Large Language Model for Finance","ref_index":44,"is_internal_anchor":true},{"citing_arxiv_id":"2604.24037","citing_title":"A Limit Theory of Foundation Models: A Mathematical Approach to Understanding Emergent Intelligence and Scaling Laws","ref_index":65,"is_internal_anchor":true},{"citing_arxiv_id":"2605.09189","citing_title":"Practical Scaling Laws: Converting Compute into Performance in a Data-Constrained World","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08297","citing_title":"A Qualitative Test-Risk Mechanism for Scaling Behavior in Normalized Residual Networks","ref_index":2,"is_internal_anchor":true},{"citing_arxiv_id":"2511.16624","citing_title":"SAM 3D: 3Dfy Anything in Images","ref_index":15,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07546","citing_title":"On the Invariance and Generality of Neural Scaling Laws","ref_index":26,"is_internal_anchor":true},{"citing_arxiv_id":"2605.06859","citing_title":"Knowledge Transfer Scaling Laws for 3D Medical Imaging","ref_index":15,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07302","citing_title":"Pretraining Induces a Reusable Spectral Basis for Downstream Task Adaptation","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2207.05221","citing_title":"Language Models (Mostly) Know What They Know","ref_index":66,"is_internal_anchor":true},{"citing_arxiv_id":"2204.05862","citing_title":"Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback","ref_index":9,"is_internal_anchor":true},{"citing_arxiv_id":"2604.17267","citing_title":"Rectification Difficulty and Optimal Sample Allocation in LLM-Augmented Surveys","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2605.02853","citing_title":"Trust, but Verify: Peeling Low-Bit Transformer Networks for Training Monitoring","ref_index":8,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EHVUIWH5KI7W26ODTL3DG5DFCK","json":"https://pith.science/pith/EHVUIWH5KI7W26ODTL3DG5DFCK.json","graph_json":"https://pith.science/api/pith-number/EHVUIWH5KI7W26ODTL3DG5DFCK/graph.json","events_json":"https://pith.science/api/pith-number/EHVUIWH5KI7W26ODTL3DG5DFCK/events.json","paper":"https://pith.science/paper/EHVUIWH5"},"agent_actions":{"view_html":"https://pith.science/pith/EHVUIWH5KI7W26ODTL3DG5DFCK","download_json":"https://pith.science/pith/EHVUIWH5KI7W26ODTL3DG5DFCK.json","view_paper":"https://pith.science/paper/EHVUIWH5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2102.01293&json=true","fetch_graph":"https://pith.science/api/pith-number/EHVUIWH5KI7W26ODTL3DG5DFCK/graph.json","fetch_events":"https://pith.science/api/pith-number/EHVUIWH5KI7W26ODTL3DG5DFCK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EHVUIWH5KI7W26ODTL3DG5DFCK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EHVUIWH5KI7W26ODTL3DG5DFCK/action/storage_attestation","attest_author":"https://pith.science/pith/EHVUIWH5KI7W26ODTL3DG5DFCK/action/author_attestation","sign_citation":"https://pith.science/pith/EHVUIWH5KI7W26ODTL3DG5DFCK/action/citation_signature","submit_replication":"https://pith.science/pith/EHVUIWH5KI7W26ODTL3DG5DFCK/action/replication_record"}},"created_at":"2026-05-18T00:52:06.250826+00:00","updated_at":"2026-05-18T00:52:06.250826+00:00"}