{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:QYEAEV557T2HNPNJQSTBBNFS25","short_pith_number":"pith:QYEAEV55","schema_version":"1.0","canonical_sha256":"86080257bdfcf476bda984a610b4b2d753afa7258192caddf292e5c0820e0d5c","source":{"kind":"arxiv","id":"2003.01897","version":2},"attestation_state":"computed","paper":{"title":"Optimal Regularization Can Mitigate Double Descent","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NE","math.ST","stat.ML","stat.TH"],"primary_cat":"cs.LG","authors_text":"Prayaag Venkat, Preetum Nakkiran, Sham Kakade, Tengyu Ma","submitted_at":"2020-03-04T05:19:09Z","abstract_excerpt":"Recent empirical and theoretical studies have shown that many learning algorithms -- from linear regression to neural networks -- can have test performance that is non-monotonic in quantities such the sample size and model size. This striking phenomenon, often referred to as \"double descent\", has raised questions of if we need to re-think our current understanding of generalization. In this work, we study whether the double-descent phenomenon can be avoided by using optimal regularization. Theoretically, we prove that for certain linear regression models with isotropic data distribution, optim"},"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":"2003.01897","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-03-04T05:19:09Z","cross_cats_sorted":["cs.NE","math.ST","stat.ML","stat.TH"],"title_canon_sha256":"4c4d89a834c5fa8336586d746141cb2891e58e8bec63b0d99aba7f8ac12c6608","abstract_canon_sha256":"ad46365eaff9fceae7aad9b35af6ac1a2e42e10e71b503818d97ae2296233363"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:36:06.891612Z","signature_b64":"xL7AcO6OPGjVPtkmSDgIGKqJnyN/BNnyXGAxNvOuMvtCTp2xl+6Ew+cv6AeafMuVjVsx2IaVJbUWFjFmD0s1BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"86080257bdfcf476bda984a610b4b2d753afa7258192caddf292e5c0820e0d5c","last_reissued_at":"2026-07-05T02:36:06.891179Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:36:06.891179Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Optimal Regularization Can Mitigate Double Descent","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NE","math.ST","stat.ML","stat.TH"],"primary_cat":"cs.LG","authors_text":"Prayaag Venkat, Preetum Nakkiran, Sham Kakade, Tengyu Ma","submitted_at":"2020-03-04T05:19:09Z","abstract_excerpt":"Recent empirical and theoretical studies have shown that many learning algorithms -- from linear regression to neural networks -- can have test performance that is non-monotonic in quantities such the sample size and model size. This striking phenomenon, often referred to as \"double descent\", has raised questions of if we need to re-think our current understanding of generalization. In this work, we study whether the double-descent phenomenon can be avoided by using optimal regularization. Theoretically, we prove that for certain linear regression models with isotropic data distribution, optim"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2003.01897","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/2003.01897/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":"2003.01897","created_at":"2026-07-05T02:36:06.891243+00:00"},{"alias_kind":"arxiv_version","alias_value":"2003.01897v2","created_at":"2026-07-05T02:36:06.891243+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2003.01897","created_at":"2026-07-05T02:36:06.891243+00:00"},{"alias_kind":"pith_short_12","alias_value":"QYEAEV557T2H","created_at":"2026-07-05T02:36:06.891243+00:00"},{"alias_kind":"pith_short_16","alias_value":"QYEAEV557T2HNPNJ","created_at":"2026-07-05T02:36:06.891243+00:00"},{"alias_kind":"pith_short_8","alias_value":"QYEAEV55","created_at":"2026-07-05T02:36:06.891243+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.10089","citing_title":"A Theory on Flow Matching with Neural Networks","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31110","citing_title":"Explaining Machine Learning and Memorization with Statistical Mechanics","ref_index":162,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18022","citing_title":"Unveiling Memorization-Generalization Coexistence: A Case Study on Arithmetic Tasks with Label Noise","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2506.07298","citing_title":"Pre-trained Large Language Models Learn Hidden Markov Models In-context","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2508.17412","citing_title":"A Ridge Too Far: Correcting Over-Shrinkage via Negative Regularization","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17202","citing_title":"Double Descent in Quantum Kernel Ridge Regression","ref_index":18,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QYEAEV557T2HNPNJQSTBBNFS25","json":"https://pith.science/pith/QYEAEV557T2HNPNJQSTBBNFS25.json","graph_json":"https://pith.science/api/pith-number/QYEAEV557T2HNPNJQSTBBNFS25/graph.json","events_json":"https://pith.science/api/pith-number/QYEAEV557T2HNPNJQSTBBNFS25/events.json","paper":"https://pith.science/paper/QYEAEV55"},"agent_actions":{"view_html":"https://pith.science/pith/QYEAEV557T2HNPNJQSTBBNFS25","download_json":"https://pith.science/pith/QYEAEV557T2HNPNJQSTBBNFS25.json","view_paper":"https://pith.science/paper/QYEAEV55","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2003.01897&json=true","fetch_graph":"https://pith.science/api/pith-number/QYEAEV557T2HNPNJQSTBBNFS25/graph.json","fetch_events":"https://pith.science/api/pith-number/QYEAEV557T2HNPNJQSTBBNFS25/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QYEAEV557T2HNPNJQSTBBNFS25/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QYEAEV557T2HNPNJQSTBBNFS25/action/storage_attestation","attest_author":"https://pith.science/pith/QYEAEV557T2HNPNJQSTBBNFS25/action/author_attestation","sign_citation":"https://pith.science/pith/QYEAEV557T2HNPNJQSTBBNFS25/action/citation_signature","submit_replication":"https://pith.science/pith/QYEAEV557T2HNPNJQSTBBNFS25/action/replication_record"}},"created_at":"2026-07-05T02:36:06.891243+00:00","updated_at":"2026-07-05T02:36:06.891243+00:00"}