{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:DJI7WUEEFFLAZ7TXDNAMUFB5YW","short_pith_number":"pith:DJI7WUEE","schema_version":"1.0","canonical_sha256":"1a51fb508429560cfe771b40ca143dc59f89e1e1eae16e9bd4cf387fcf6a00f7","source":{"kind":"arxiv","id":"2410.19092","version":1},"attestation_state":"computed","paper":{"title":"Provable Tempered Overfitting of Minimal Nets and Typical Nets","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Daniel Soudry, Gal Vardi, Itamar Harel, Itay Evron, Nathan Srebro, William M. Hoza","submitted_at":"2024-10-24T18:51:56Z","abstract_excerpt":"We study the overfitting behavior of fully connected deep Neural Networks (NNs) with binary weights fitted to perfectly classify a noisy training set. We consider interpolation using both the smallest NN (having the minimal number of weights) and a random interpolating NN. For both learning rules, we prove overfitting is tempered. Our analysis rests on a new bound on the size of a threshold circuit consistent with a partial function. To the best of our knowledge, ours are the first theoretical results on benign or tempered overfitting that: (1) apply to deep NNs, and (2) do not require a very "},"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":"2410.19092","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-24T18:51:56Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"ee8c893e2a04ac40b91a386a5fa7c68c8d62d2cda3fbe3d0e5338327de3743cd","abstract_canon_sha256":"48fa43e007c68456dcec22ea37bd48bc493c56f22dd04fb8d935eedbf821811f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:25:51.324813Z","signature_b64":"k50j8PyZpYingvYvbFXJBcK8iFVvZSnqPcfO7MdQZKkTkcuajnp0QjTWxwrPSd8nbHkumUzLIiUvcj+k49HyBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1a51fb508429560cfe771b40ca143dc59f89e1e1eae16e9bd4cf387fcf6a00f7","last_reissued_at":"2026-07-05T09:25:51.324313Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:25:51.324313Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Provable Tempered Overfitting of Minimal Nets and Typical Nets","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Daniel Soudry, Gal Vardi, Itamar Harel, Itay Evron, Nathan Srebro, William M. Hoza","submitted_at":"2024-10-24T18:51:56Z","abstract_excerpt":"We study the overfitting behavior of fully connected deep Neural Networks (NNs) with binary weights fitted to perfectly classify a noisy training set. We consider interpolation using both the smallest NN (having the minimal number of weights) and a random interpolating NN. For both learning rules, we prove overfitting is tempered. Our analysis rests on a new bound on the size of a threshold circuit consistent with a partial function. To the best of our knowledge, ours are the first theoretical results on benign or tempered overfitting that: (1) apply to deep NNs, and (2) do not require a very "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.19092","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/2410.19092/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":"2410.19092","created_at":"2026-07-05T09:25:51.324382+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.19092v1","created_at":"2026-07-05T09:25:51.324382+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.19092","created_at":"2026-07-05T09:25:51.324382+00:00"},{"alias_kind":"pith_short_12","alias_value":"DJI7WUEEFFLA","created_at":"2026-07-05T09:25:51.324382+00:00"},{"alias_kind":"pith_short_16","alias_value":"DJI7WUEEFFLAZ7TX","created_at":"2026-07-05T09:25:51.324382+00:00"},{"alias_kind":"pith_short_8","alias_value":"DJI7WUEE","created_at":"2026-07-05T09:25:51.324382+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.11621","citing_title":"A Classical View on Benign Overfitting: The Role of Sample Size","ref_index":35,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DJI7WUEEFFLAZ7TXDNAMUFB5YW","json":"https://pith.science/pith/DJI7WUEEFFLAZ7TXDNAMUFB5YW.json","graph_json":"https://pith.science/api/pith-number/DJI7WUEEFFLAZ7TXDNAMUFB5YW/graph.json","events_json":"https://pith.science/api/pith-number/DJI7WUEEFFLAZ7TXDNAMUFB5YW/events.json","paper":"https://pith.science/paper/DJI7WUEE"},"agent_actions":{"view_html":"https://pith.science/pith/DJI7WUEEFFLAZ7TXDNAMUFB5YW","download_json":"https://pith.science/pith/DJI7WUEEFFLAZ7TXDNAMUFB5YW.json","view_paper":"https://pith.science/paper/DJI7WUEE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.19092&json=true","fetch_graph":"https://pith.science/api/pith-number/DJI7WUEEFFLAZ7TXDNAMUFB5YW/graph.json","fetch_events":"https://pith.science/api/pith-number/DJI7WUEEFFLAZ7TXDNAMUFB5YW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DJI7WUEEFFLAZ7TXDNAMUFB5YW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DJI7WUEEFFLAZ7TXDNAMUFB5YW/action/storage_attestation","attest_author":"https://pith.science/pith/DJI7WUEEFFLAZ7TXDNAMUFB5YW/action/author_attestation","sign_citation":"https://pith.science/pith/DJI7WUEEFFLAZ7TXDNAMUFB5YW/action/citation_signature","submit_replication":"https://pith.science/pith/DJI7WUEEFFLAZ7TXDNAMUFB5YW/action/replication_record"}},"created_at":"2026-07-05T09:25:51.324382+00:00","updated_at":"2026-07-05T09:25:51.324382+00:00"}