{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:YJEHQWKL7FOEZHF5GAWJLW7YEQ","short_pith_number":"pith:YJEHQWKL","schema_version":"1.0","canonical_sha256":"c24878594bf95c4c9cbd302c95dbf82404aebdc3c8e50ed3f6fc2e316dd662ac","source":{"kind":"arxiv","id":"2410.16073","version":1},"attestation_state":"computed","paper":{"title":"On the Geometry of Regularization in Adversarial Training: High-Dimensional Asymptotics and Generalization Bounds","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cond-mat.dis-nn","cs.LG","math.ST","stat.TH"],"primary_cat":"stat.ML","authors_text":"Bruno Loureiro, Julia Kempe, Matteo Vilucchio, Nikolaos Tsilivis","submitted_at":"2024-10-21T14:53:12Z","abstract_excerpt":"Regularization, whether explicit in terms of a penalty in the loss or implicit in the choice of algorithm, is a cornerstone of modern machine learning. Indeed, controlling the complexity of the model class is particularly important when data is scarce, noisy or contaminated, as it translates a statistical belief on the underlying structure of the data. This work investigates the question of how to choose the regularization norm $\\lVert \\cdot \\rVert$ in the context of high-dimensional adversarial training for binary classification. To this end, we first derive an exact asymptotic description of"},"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":"2410.16073","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2024-10-21T14:53:12Z","cross_cats_sorted":["cond-mat.dis-nn","cs.LG","math.ST","stat.TH"],"title_canon_sha256":"76900b3b1de4fbfa6f28169f1af8c13cbbac48772b27d27d3b40058e96a7db93","abstract_canon_sha256":"f0359cad84b36210dd8d0607214b748125bdf85104eb58c44edbfad9aae8e3c9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:23:28.939676Z","signature_b64":"w1aReQ0e1P2zTOo9HTSB/re2Gj0nuPLP/M2KMpguuVeizqxreezSn/w01BW5YP74yKMc6XqwYsBMuTBFHoJCCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c24878594bf95c4c9cbd302c95dbf82404aebdc3c8e50ed3f6fc2e316dd662ac","last_reissued_at":"2026-07-05T09:23:28.939143Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:23:28.939143Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On the Geometry of Regularization in Adversarial Training: High-Dimensional Asymptotics and Generalization Bounds","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cond-mat.dis-nn","cs.LG","math.ST","stat.TH"],"primary_cat":"stat.ML","authors_text":"Bruno Loureiro, Julia Kempe, Matteo Vilucchio, Nikolaos Tsilivis","submitted_at":"2024-10-21T14:53:12Z","abstract_excerpt":"Regularization, whether explicit in terms of a penalty in the loss or implicit in the choice of algorithm, is a cornerstone of modern machine learning. Indeed, controlling the complexity of the model class is particularly important when data is scarce, noisy or contaminated, as it translates a statistical belief on the underlying structure of the data. This work investigates the question of how to choose the regularization norm $\\lVert \\cdot \\rVert$ in the context of high-dimensional adversarial training for binary classification. To this end, we first derive an exact asymptotic description of"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.16073","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/2410.16073/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":"2410.16073","created_at":"2026-07-05T09:23:28.939213+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.16073v1","created_at":"2026-07-05T09:23:28.939213+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.16073","created_at":"2026-07-05T09:23:28.939213+00:00"},{"alias_kind":"pith_short_12","alias_value":"YJEHQWKL7FOE","created_at":"2026-07-05T09:23:28.939213+00:00"},{"alias_kind":"pith_short_16","alias_value":"YJEHQWKL7FOEZHF5","created_at":"2026-07-05T09:23:28.939213+00:00"},{"alias_kind":"pith_short_8","alias_value":"YJEHQWKL","created_at":"2026-07-05T09:23:28.939213+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.00207","citing_title":"Homogenization of $\\ell_2$-Adversarial Training in High-Dimensions: Exact Dynamics under Stochastic Gradient Descent","ref_index":66,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31110","citing_title":"Explaining Machine Learning and Memorization with Statistical Mechanics","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YJEHQWKL7FOEZHF5GAWJLW7YEQ","json":"https://pith.science/pith/YJEHQWKL7FOEZHF5GAWJLW7YEQ.json","graph_json":"https://pith.science/api/pith-number/YJEHQWKL7FOEZHF5GAWJLW7YEQ/graph.json","events_json":"https://pith.science/api/pith-number/YJEHQWKL7FOEZHF5GAWJLW7YEQ/events.json","paper":"https://pith.science/paper/YJEHQWKL"},"agent_actions":{"view_html":"https://pith.science/pith/YJEHQWKL7FOEZHF5GAWJLW7YEQ","download_json":"https://pith.science/pith/YJEHQWKL7FOEZHF5GAWJLW7YEQ.json","view_paper":"https://pith.science/paper/YJEHQWKL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.16073&json=true","fetch_graph":"https://pith.science/api/pith-number/YJEHQWKL7FOEZHF5GAWJLW7YEQ/graph.json","fetch_events":"https://pith.science/api/pith-number/YJEHQWKL7FOEZHF5GAWJLW7YEQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YJEHQWKL7FOEZHF5GAWJLW7YEQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YJEHQWKL7FOEZHF5GAWJLW7YEQ/action/storage_attestation","attest_author":"https://pith.science/pith/YJEHQWKL7FOEZHF5GAWJLW7YEQ/action/author_attestation","sign_citation":"https://pith.science/pith/YJEHQWKL7FOEZHF5GAWJLW7YEQ/action/citation_signature","submit_replication":"https://pith.science/pith/YJEHQWKL7FOEZHF5GAWJLW7YEQ/action/replication_record"}},"created_at":"2026-07-05T09:23:28.939213+00:00","updated_at":"2026-07-05T09:23:28.939213+00:00"}