{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:B7MLTWPVTNPKN3KNXMPAI6IBDY","short_pith_number":"pith:B7MLTWPV","schema_version":"1.0","canonical_sha256":"0fd8b9d9f59b5ea6ed4dbb1e0479011e10311977f3c3474fcc5b7d10be7c2601","source":{"kind":"arxiv","id":"2107.09550","version":2},"attestation_state":"computed","paper":{"title":"Convergence rates for shallow neural networks learned by gradient descent","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.TH"],"primary_cat":"math.ST","authors_text":"Alina Braun, Harro Walk, Michael Kohler, Sophie Langer","submitted_at":"2021-07-20T15:10:03Z","abstract_excerpt":"In this paper we analyze the $L_2$ error of neural network regression estimates with one hidden layer. Under the assumption that the Fourier transform of the regression function decays suitably fast, we show that an estimate, where all initial weights are chosen according to proper uniform distributions and where the weights are learned by gradient descent, achieves a rate of convergence of $1/\\sqrt{n}$ (up to a logarithmic factor). Our statistical analysis implies that the key aspect behind this result is the proper choice of the initial inner weights and the adjustment of the outer weights v"},"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":"2107.09550","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"math.ST","submitted_at":"2021-07-20T15:10:03Z","cross_cats_sorted":["stat.TH"],"title_canon_sha256":"a7b09de511c50e3ed1296d44e4158ef1c5fe8a4a9609e31d0f69a36da69d69db","abstract_canon_sha256":"c521431331dec8b29d13c69867fa7027a7b4867156a94abfcdeddbd1d2b373de"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:42:15.263354Z","signature_b64":"KaRQa+uWjngb98TKINFdpS59GAAaOfS7AdODpxGpHRpAEEHYaSlTfguSbMROZ407dmKObtT1gLVdsLDcaVomCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0fd8b9d9f59b5ea6ed4dbb1e0479011e10311977f3c3474fcc5b7d10be7c2601","last_reissued_at":"2026-07-05T06:42:15.262900Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:42:15.262900Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Convergence rates for shallow neural networks learned by gradient descent","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.TH"],"primary_cat":"math.ST","authors_text":"Alina Braun, Harro Walk, Michael Kohler, Sophie Langer","submitted_at":"2021-07-20T15:10:03Z","abstract_excerpt":"In this paper we analyze the $L_2$ error of neural network regression estimates with one hidden layer. Under the assumption that the Fourier transform of the regression function decays suitably fast, we show that an estimate, where all initial weights are chosen according to proper uniform distributions and where the weights are learned by gradient descent, achieves a rate of convergence of $1/\\sqrt{n}$ (up to a logarithmic factor). Our statistical analysis implies that the key aspect behind this result is the proper choice of the initial inner weights and the adjustment of the outer weights v"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2107.09550","kind":"arxiv","version":2},"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/2107.09550/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":"2107.09550","created_at":"2026-07-05T06:42:15.262955+00:00"},{"alias_kind":"arxiv_version","alias_value":"2107.09550v2","created_at":"2026-07-05T06:42:15.262955+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2107.09550","created_at":"2026-07-05T06:42:15.262955+00:00"},{"alias_kind":"pith_short_12","alias_value":"B7MLTWPVTNPK","created_at":"2026-07-05T06:42:15.262955+00:00"},{"alias_kind":"pith_short_16","alias_value":"B7MLTWPVTNPKN3KN","created_at":"2026-07-05T06:42:15.262955+00:00"},{"alias_kind":"pith_short_8","alias_value":"B7MLTWPV","created_at":"2026-07-05T06:42:15.262955+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/B7MLTWPVTNPKN3KNXMPAI6IBDY","json":"https://pith.science/pith/B7MLTWPVTNPKN3KNXMPAI6IBDY.json","graph_json":"https://pith.science/api/pith-number/B7MLTWPVTNPKN3KNXMPAI6IBDY/graph.json","events_json":"https://pith.science/api/pith-number/B7MLTWPVTNPKN3KNXMPAI6IBDY/events.json","paper":"https://pith.science/paper/B7MLTWPV"},"agent_actions":{"view_html":"https://pith.science/pith/B7MLTWPVTNPKN3KNXMPAI6IBDY","download_json":"https://pith.science/pith/B7MLTWPVTNPKN3KNXMPAI6IBDY.json","view_paper":"https://pith.science/paper/B7MLTWPV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2107.09550&json=true","fetch_graph":"https://pith.science/api/pith-number/B7MLTWPVTNPKN3KNXMPAI6IBDY/graph.json","fetch_events":"https://pith.science/api/pith-number/B7MLTWPVTNPKN3KNXMPAI6IBDY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/B7MLTWPVTNPKN3KNXMPAI6IBDY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/B7MLTWPVTNPKN3KNXMPAI6IBDY/action/storage_attestation","attest_author":"https://pith.science/pith/B7MLTWPVTNPKN3KNXMPAI6IBDY/action/author_attestation","sign_citation":"https://pith.science/pith/B7MLTWPVTNPKN3KNXMPAI6IBDY/action/citation_signature","submit_replication":"https://pith.science/pith/B7MLTWPVTNPKN3KNXMPAI6IBDY/action/replication_record"}},"created_at":"2026-07-05T06:42:15.262955+00:00","updated_at":"2026-07-05T06:42:15.262955+00:00"}