{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ED66CFUDUGYNHJYYYCSAYFEFGP","short_pith_number":"pith:ED66CFUD","schema_version":"1.0","canonical_sha256":"20fde11683a1b0d3a718c0a40c148533e51441da54253fc37907a12847e7c724","source":{"kind":"arxiv","id":"2607.00207","version":1},"attestation_state":"computed","paper":{"title":"Homogenization of $\\ell_2$-Adversarial Training in High-Dimensions: Exact Dynamics under Stochastic Gradient Descent","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","math.PR","stat.ML"],"primary_cat":"math.OC","authors_text":"Fabrizzio Sabelli","submitted_at":"2026-06-30T21:38:10Z","abstract_excerpt":"We develop a framework for analyzing the learning dynamics of $\\ell_2$-adversarial training of single-index models on Gaussian mixtures in the high-dimensional limit under streaming stochastic gradient descent (SGD). We derive deterministic equivalents for a broad class of statistics of the SGD iterates, including the adversarial risk and distance to adversarial optimality, in terms of the solution to a system of ODEs. We use them to study two idealized learning rate schedules: the Polyak stepsize and exact line search. In the case of $\\ell_2$-adversarial least squares with a single class, we "},"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":"2607.00207","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"math.OC","submitted_at":"2026-06-30T21:38:10Z","cross_cats_sorted":["cs.LG","math.PR","stat.ML"],"title_canon_sha256":"42ddfb30d942a2dd57494cf63dd99f253724e2c826ff90a2f99f2d32336a38a6","abstract_canon_sha256":"89f0ae6d3130b9356caf7a8b1e40df24bc49896ce5caa7d9e4ebb5d6e5bd523c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-02T00:18:39.078132Z","signature_b64":"0TrG7D1RAk5/klj9wRF8jJEFabP9lbsEtcA6nOD1+9nzhK27PABJkyPQOgfQcJWDu0xfqFWHWirp9WHIB3SmAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"20fde11683a1b0d3a718c0a40c148533e51441da54253fc37907a12847e7c724","last_reissued_at":"2026-07-02T00:18:39.077478Z","signature_status":"signed_v1","first_computed_at":"2026-07-02T00:18:39.077478Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Homogenization of $\\ell_2$-Adversarial Training in High-Dimensions: Exact Dynamics under Stochastic Gradient Descent","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","math.PR","stat.ML"],"primary_cat":"math.OC","authors_text":"Fabrizzio Sabelli","submitted_at":"2026-06-30T21:38:10Z","abstract_excerpt":"We develop a framework for analyzing the learning dynamics of $\\ell_2$-adversarial training of single-index models on Gaussian mixtures in the high-dimensional limit under streaming stochastic gradient descent (SGD). We derive deterministic equivalents for a broad class of statistics of the SGD iterates, including the adversarial risk and distance to adversarial optimality, in terms of the solution to a system of ODEs. We use them to study two idealized learning rate schedules: the Polyak stepsize and exact line search. In the case of $\\ell_2$-adversarial least squares with a single class, we "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.00207","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/2607.00207/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":"2607.00207","created_at":"2026-07-02T00:18:39.077582+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.00207v1","created_at":"2026-07-02T00:18:39.077582+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.00207","created_at":"2026-07-02T00:18:39.077582+00:00"},{"alias_kind":"pith_short_12","alias_value":"ED66CFUDUGYN","created_at":"2026-07-02T00:18:39.077582+00:00"},{"alias_kind":"pith_short_16","alias_value":"ED66CFUDUGYNHJYY","created_at":"2026-07-02T00:18:39.077582+00:00"},{"alias_kind":"pith_short_8","alias_value":"ED66CFUD","created_at":"2026-07-02T00:18:39.077582+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/ED66CFUDUGYNHJYYYCSAYFEFGP","json":"https://pith.science/pith/ED66CFUDUGYNHJYYYCSAYFEFGP.json","graph_json":"https://pith.science/api/pith-number/ED66CFUDUGYNHJYYYCSAYFEFGP/graph.json","events_json":"https://pith.science/api/pith-number/ED66CFUDUGYNHJYYYCSAYFEFGP/events.json","paper":"https://pith.science/paper/ED66CFUD"},"agent_actions":{"view_html":"https://pith.science/pith/ED66CFUDUGYNHJYYYCSAYFEFGP","download_json":"https://pith.science/pith/ED66CFUDUGYNHJYYYCSAYFEFGP.json","view_paper":"https://pith.science/paper/ED66CFUD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.00207&json=true","fetch_graph":"https://pith.science/api/pith-number/ED66CFUDUGYNHJYYYCSAYFEFGP/graph.json","fetch_events":"https://pith.science/api/pith-number/ED66CFUDUGYNHJYYYCSAYFEFGP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ED66CFUDUGYNHJYYYCSAYFEFGP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ED66CFUDUGYNHJYYYCSAYFEFGP/action/storage_attestation","attest_author":"https://pith.science/pith/ED66CFUDUGYNHJYYYCSAYFEFGP/action/author_attestation","sign_citation":"https://pith.science/pith/ED66CFUDUGYNHJYYYCSAYFEFGP/action/citation_signature","submit_replication":"https://pith.science/pith/ED66CFUDUGYNHJYYYCSAYFEFGP/action/replication_record"}},"created_at":"2026-07-02T00:18:39.077582+00:00","updated_at":"2026-07-02T00:18:39.077582+00:00"}