{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:G325O3YSZMFXWVNNTAMHWRXS2D","short_pith_number":"pith:G325O3YS","schema_version":"1.0","canonical_sha256":"36f5d76f12cb0b7b55ad98187b46f2d0ca548abacc5867ef3b076a921244de56","source":{"kind":"arxiv","id":"2412.07942","version":1},"attestation_state":"computed","paper":{"title":"Neural Scaling Laws Rooted in the Data Distribution","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cond-mat.dis-nn"],"primary_cat":"cs.LG","authors_text":"Ari Brill","submitted_at":"2024-12-10T22:01:38Z","abstract_excerpt":"Deep neural networks exhibit empirical neural scaling laws, with error decreasing as a power law with increasing model or data size, across a wide variety of architectures, tasks, and datasets. This universality suggests that scaling laws may result from general properties of natural learning tasks. We develop a mathematical model intended to describe natural datasets using percolation theory. Two distinct criticality regimes emerge, each yielding optimal power-law neural scaling laws. These regimes, corresponding to power-law-distributed discrete subtasks and a dominant data manifold, can be "},"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":"2412.07942","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-10T22:01:38Z","cross_cats_sorted":["cond-mat.dis-nn"],"title_canon_sha256":"fd0336e06e8ec70691aa52a37aeb89c6b853bfc6772b2729b2eb095f1104bfd8","abstract_canon_sha256":"060dd13d3f1bcf3294f05bb561bfc4649dc3a88ae123b4273052978d9042bbbd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:47:41.500851Z","signature_b64":"0e5POKW2BwB/J/eR8iElAe+ohDoUSuYers+34YJP6RxZBt5QStGMddT4qwEBhdzCG/ZOtMGmyJQV3wZcJK1vBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"36f5d76f12cb0b7b55ad98187b46f2d0ca548abacc5867ef3b076a921244de56","last_reissued_at":"2026-07-05T09:47:41.500387Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:47:41.500387Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Neural Scaling Laws Rooted in the Data Distribution","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cond-mat.dis-nn"],"primary_cat":"cs.LG","authors_text":"Ari Brill","submitted_at":"2024-12-10T22:01:38Z","abstract_excerpt":"Deep neural networks exhibit empirical neural scaling laws, with error decreasing as a power law with increasing model or data size, across a wide variety of architectures, tasks, and datasets. This universality suggests that scaling laws may result from general properties of natural learning tasks. We develop a mathematical model intended to describe natural datasets using percolation theory. Two distinct criticality regimes emerge, each yielding optimal power-law neural scaling laws. These regimes, corresponding to power-law-distributed discrete subtasks and a dominant data manifold, can be "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.07942","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/2412.07942/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":"2412.07942","created_at":"2026-07-05T09:47:41.500446+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.07942v1","created_at":"2026-07-05T09:47:41.500446+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.07942","created_at":"2026-07-05T09:47:41.500446+00:00"},{"alias_kind":"pith_short_12","alias_value":"G325O3YSZMFX","created_at":"2026-07-05T09:47:41.500446+00:00"},{"alias_kind":"pith_short_16","alias_value":"G325O3YSZMFXWVNN","created_at":"2026-07-05T09:47:41.500446+00:00"},{"alias_kind":"pith_short_8","alias_value":"G325O3YS","created_at":"2026-07-05T09:47:41.500446+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20347","citing_title":"Critical Percolation as a Synthetic Data Model for Interpretability","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29858","citing_title":"Smooth Scaling Laws Hide Stepwise Token Learning","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09189","citing_title":"Practical Scaling Laws: Converting Compute into Performance in a Data-Constrained World","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2505.10465","citing_title":"Superposition Yields Robust Neural Scaling","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/G325O3YSZMFXWVNNTAMHWRXS2D","json":"https://pith.science/pith/G325O3YSZMFXWVNNTAMHWRXS2D.json","graph_json":"https://pith.science/api/pith-number/G325O3YSZMFXWVNNTAMHWRXS2D/graph.json","events_json":"https://pith.science/api/pith-number/G325O3YSZMFXWVNNTAMHWRXS2D/events.json","paper":"https://pith.science/paper/G325O3YS"},"agent_actions":{"view_html":"https://pith.science/pith/G325O3YSZMFXWVNNTAMHWRXS2D","download_json":"https://pith.science/pith/G325O3YSZMFXWVNNTAMHWRXS2D.json","view_paper":"https://pith.science/paper/G325O3YS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.07942&json=true","fetch_graph":"https://pith.science/api/pith-number/G325O3YSZMFXWVNNTAMHWRXS2D/graph.json","fetch_events":"https://pith.science/api/pith-number/G325O3YSZMFXWVNNTAMHWRXS2D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/G325O3YSZMFXWVNNTAMHWRXS2D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/G325O3YSZMFXWVNNTAMHWRXS2D/action/storage_attestation","attest_author":"https://pith.science/pith/G325O3YSZMFXWVNNTAMHWRXS2D/action/author_attestation","sign_citation":"https://pith.science/pith/G325O3YSZMFXWVNNTAMHWRXS2D/action/citation_signature","submit_replication":"https://pith.science/pith/G325O3YSZMFXWVNNTAMHWRXS2D/action/replication_record"}},"created_at":"2026-07-05T09:47:41.500446+00:00","updated_at":"2026-07-05T09:47:41.500446+00:00"}