{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:H2D4BCQE6TVTEQEV3YGICE7N2L","short_pith_number":"pith:H2D4BCQE","schema_version":"1.0","canonical_sha256":"3e87c08a04f4eb324095de0c8113edd2ce81ec661e01563931f9623e518f1d75","source":{"kind":"arxiv","id":"2304.03408","version":3},"attestation_state":"computed","paper":{"title":"Dynamics of Finite Width Kernel and Prediction Fluctuations in Mean Field Neural Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cond-mat.dis-nn","cs.LG"],"primary_cat":"stat.ML","authors_text":"Blake Bordelon, Cengiz Pehlevan","submitted_at":"2023-04-06T23:11:49Z","abstract_excerpt":"We analyze the dynamics of finite width effects in wide but finite feature learning neural networks. Starting from a dynamical mean field theory description of infinite width deep neural network kernel and prediction dynamics, we provide a characterization of the $O(1/\\sqrt{\\text{width}})$ fluctuations of the DMFT order parameters over random initializations of the network weights. Our results, while perturbative in width, unlike prior analyses, are non-perturbative in the strength of feature learning. In the lazy limit of network training, all kernels are random but static in time and the pre"},"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":"2304.03408","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2023-04-06T23:11:49Z","cross_cats_sorted":["cond-mat.dis-nn","cs.LG"],"title_canon_sha256":"cc3b4c9c5c1d631bce50a7c67260e90971b6d34d0a5e8b67dd64776205c55352","abstract_canon_sha256":"33d5a7346d5fc3772c54c9a44c805a3f0d9d304cc7e497890d93b37f74d5a320"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:09:47.043468Z","signature_b64":"5gmgljinFidAlw5PtOh4cvzhZXe9GrtSfdIizEXG6m9/usRQfILIWfR+SCiyLozamVoFAiKt6Jg8TDN6wNQACw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3e87c08a04f4eb324095de0c8113edd2ce81ec661e01563931f9623e518f1d75","last_reissued_at":"2026-07-05T07:09:47.042965Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:09:47.042965Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Dynamics of Finite Width Kernel and Prediction Fluctuations in Mean Field Neural Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cond-mat.dis-nn","cs.LG"],"primary_cat":"stat.ML","authors_text":"Blake Bordelon, Cengiz Pehlevan","submitted_at":"2023-04-06T23:11:49Z","abstract_excerpt":"We analyze the dynamics of finite width effects in wide but finite feature learning neural networks. Starting from a dynamical mean field theory description of infinite width deep neural network kernel and prediction dynamics, we provide a characterization of the $O(1/\\sqrt{\\text{width}})$ fluctuations of the DMFT order parameters over random initializations of the network weights. Our results, while perturbative in width, unlike prior analyses, are non-perturbative in the strength of feature learning. In the lazy limit of network training, all kernels are random but static in time and the pre"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2304.03408","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/2304.03408/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":"2304.03408","created_at":"2026-07-05T07:09:47.043025+00:00"},{"alias_kind":"arxiv_version","alias_value":"2304.03408v3","created_at":"2026-07-05T07:09:47.043025+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2304.03408","created_at":"2026-07-05T07:09:47.043025+00:00"},{"alias_kind":"pith_short_12","alias_value":"H2D4BCQE6TVT","created_at":"2026-07-05T07:09:47.043025+00:00"},{"alias_kind":"pith_short_16","alias_value":"H2D4BCQE6TVTEQEV","created_at":"2026-07-05T07:09:47.043025+00:00"},{"alias_kind":"pith_short_8","alias_value":"H2D4BCQE","created_at":"2026-07-05T07:09:47.043025+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.07870","citing_title":"Spectral Dynamics in Deep Networks: Feature Learning, Outlier Escape, and Learning Rate Transfer","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2603.11161","citing_title":"Algorithmic Task Capture, Computational Complexity, and Inductive Bias of Infinite Transformers","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12763","citing_title":"Center-Manifold Reduction of Learning at Bifurcations: Interference and Rich Learning in Recurrent Neural Networks","ref_index":87,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08746","citing_title":"The Global Empirical NTK: Self-Referential Bias and Dimensionality of Gradient Descent Learning","ref_index":87,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07870","citing_title":"Spectral Dynamics in Deep Networks: Feature Learning, Outlier Escape, and Learning Rate Transfer","ref_index":47,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/H2D4BCQE6TVTEQEV3YGICE7N2L","json":"https://pith.science/pith/H2D4BCQE6TVTEQEV3YGICE7N2L.json","graph_json":"https://pith.science/api/pith-number/H2D4BCQE6TVTEQEV3YGICE7N2L/graph.json","events_json":"https://pith.science/api/pith-number/H2D4BCQE6TVTEQEV3YGICE7N2L/events.json","paper":"https://pith.science/paper/H2D4BCQE"},"agent_actions":{"view_html":"https://pith.science/pith/H2D4BCQE6TVTEQEV3YGICE7N2L","download_json":"https://pith.science/pith/H2D4BCQE6TVTEQEV3YGICE7N2L.json","view_paper":"https://pith.science/paper/H2D4BCQE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2304.03408&json=true","fetch_graph":"https://pith.science/api/pith-number/H2D4BCQE6TVTEQEV3YGICE7N2L/graph.json","fetch_events":"https://pith.science/api/pith-number/H2D4BCQE6TVTEQEV3YGICE7N2L/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/H2D4BCQE6TVTEQEV3YGICE7N2L/action/timestamp_anchor","attest_storage":"https://pith.science/pith/H2D4BCQE6TVTEQEV3YGICE7N2L/action/storage_attestation","attest_author":"https://pith.science/pith/H2D4BCQE6TVTEQEV3YGICE7N2L/action/author_attestation","sign_citation":"https://pith.science/pith/H2D4BCQE6TVTEQEV3YGICE7N2L/action/citation_signature","submit_replication":"https://pith.science/pith/H2D4BCQE6TVTEQEV3YGICE7N2L/action/replication_record"}},"created_at":"2026-07-05T07:09:47.043025+00:00","updated_at":"2026-07-05T07:09:47.043025+00:00"}