{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:7MWP4KRCO2WA35WL5L7E6GBHES","short_pith_number":"pith:7MWP4KRC","schema_version":"1.0","canonical_sha256":"fb2cfe2a2276ac0df6cbeafe4f182724983c1e9ce36d813d061b9bf1f127d2fd","source":{"kind":"arxiv","id":"2309.08044","version":1},"attestation_state":"computed","paper":{"title":"How many Neurons do we need? A refined Analysis for Shallow Networks trained with Gradient Descent","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Mike Nguyen, Nicole M\\\"ucke","submitted_at":"2023-09-14T22:10:28Z","abstract_excerpt":"We analyze the generalization properties of two-layer neural networks in the neural tangent kernel (NTK) regime, trained with gradient descent (GD). For early stopped GD we derive fast rates of convergence that are known to be minimax optimal in the framework of non-parametric regression in reproducing kernel Hilbert spaces. On our way, we precisely keep track of the number of hidden neurons required for generalization and improve over existing results. We further show that the weights during training remain in a vicinity around initialization, the radius being dependent on structural assumpti"},"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":"2309.08044","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2023-09-14T22:10:28Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"49605f1d8f59d8fc534431dda0c40d7d50eb038a35476750a82ca4fa5706a450","abstract_canon_sha256":"e2b738bb9adfa403c180a8a57034103d458950b2c66b03adcf94344fe95170ef"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:51:01.128357Z","signature_b64":"JCYawGvphn3x6nLk0afPyBfrscVqiJ8VZApQqvff0L6UX+OGEBpaR9mzhIqmb9ARHjaYoZkt+MQqgsrXjmfTAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fb2cfe2a2276ac0df6cbeafe4f182724983c1e9ce36d813d061b9bf1f127d2fd","last_reissued_at":"2026-07-05T06:51:01.127950Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:51:01.127950Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How many Neurons do we need? A refined Analysis for Shallow Networks trained with Gradient Descent","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Mike Nguyen, Nicole M\\\"ucke","submitted_at":"2023-09-14T22:10:28Z","abstract_excerpt":"We analyze the generalization properties of two-layer neural networks in the neural tangent kernel (NTK) regime, trained with gradient descent (GD). For early stopped GD we derive fast rates of convergence that are known to be minimax optimal in the framework of non-parametric regression in reproducing kernel Hilbert spaces. On our way, we precisely keep track of the number of hidden neurons required for generalization and improve over existing results. We further show that the weights during training remain in a vicinity around initialization, the radius being dependent on structural assumpti"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.08044","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/2309.08044/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":"2309.08044","created_at":"2026-07-05T06:51:01.128005+00:00"},{"alias_kind":"arxiv_version","alias_value":"2309.08044v1","created_at":"2026-07-05T06:51:01.128005+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.08044","created_at":"2026-07-05T06:51:01.128005+00:00"},{"alias_kind":"pith_short_12","alias_value":"7MWP4KRCO2WA","created_at":"2026-07-05T06:51:01.128005+00:00"},{"alias_kind":"pith_short_16","alias_value":"7MWP4KRCO2WA35WL","created_at":"2026-07-05T06:51:01.128005+00:00"},{"alias_kind":"pith_short_8","alias_value":"7MWP4KRC","created_at":"2026-07-05T06:51:01.128005+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.17518","citing_title":"Optimal Convergence Rates for Neural Operators","ref_index":50,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7MWP4KRCO2WA35WL5L7E6GBHES","json":"https://pith.science/pith/7MWP4KRCO2WA35WL5L7E6GBHES.json","graph_json":"https://pith.science/api/pith-number/7MWP4KRCO2WA35WL5L7E6GBHES/graph.json","events_json":"https://pith.science/api/pith-number/7MWP4KRCO2WA35WL5L7E6GBHES/events.json","paper":"https://pith.science/paper/7MWP4KRC"},"agent_actions":{"view_html":"https://pith.science/pith/7MWP4KRCO2WA35WL5L7E6GBHES","download_json":"https://pith.science/pith/7MWP4KRCO2WA35WL5L7E6GBHES.json","view_paper":"https://pith.science/paper/7MWP4KRC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2309.08044&json=true","fetch_graph":"https://pith.science/api/pith-number/7MWP4KRCO2WA35WL5L7E6GBHES/graph.json","fetch_events":"https://pith.science/api/pith-number/7MWP4KRCO2WA35WL5L7E6GBHES/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7MWP4KRCO2WA35WL5L7E6GBHES/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7MWP4KRCO2WA35WL5L7E6GBHES/action/storage_attestation","attest_author":"https://pith.science/pith/7MWP4KRCO2WA35WL5L7E6GBHES/action/author_attestation","sign_citation":"https://pith.science/pith/7MWP4KRCO2WA35WL5L7E6GBHES/action/citation_signature","submit_replication":"https://pith.science/pith/7MWP4KRCO2WA35WL5L7E6GBHES/action/replication_record"}},"created_at":"2026-07-05T06:51:01.128005+00:00","updated_at":"2026-07-05T06:51:01.128005+00:00"}