{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:3HT3J3HACOCYNLL4DMKNAVTIUP","short_pith_number":"pith:3HT3J3HA","schema_version":"1.0","canonical_sha256":"d9e7b4ece0138586ad7c1b14d05668a3cac068eec6870f7ce7106bb4a4a010a9","source":{"kind":"arxiv","id":"2511.01126","version":2},"attestation_state":"computed","paper":{"title":"Stochastic Regret Guarantees for Online Zeroth- and First-Order Bilevel Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.NA","math.NA","math.OC","math.ST","stat.TH"],"primary_cat":"cs.LG","authors_text":"Bojian Hou, Davoud Ataee Tarzanagh, George Michailidis, Li Shen, Parvin Nazari","submitted_at":"2025-11-03T00:29:36Z","abstract_excerpt":"Online bilevel optimization (OBO) is a powerful framework for machine learning problems where both outer and inner objectives evolve over time, requiring dynamic updates. Current OBO approaches rely on deterministic \\textit{window-smoothed} regret minimization, which may not accurately reflect system performance when functions change rapidly. In this work, we introduce a novel search direction and show that both first- and zeroth-order (ZO) stochastic OBO algorithms leveraging this direction achieve sublinear {stochastic bilevel regret without window smoothing}. Beyond these guarantees, our fr"},"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":"2511.01126","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-03T00:29:36Z","cross_cats_sorted":["cs.NA","math.NA","math.OC","math.ST","stat.TH"],"title_canon_sha256":"a54ed121071e0b163e65bf959c3a9945b130da1c37c97686f50b840d8ecd0745","abstract_canon_sha256":"414c5c6ac553e947d66170021434eb03aa65c219087729736b3bf489802b447a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:04:16.614996Z","signature_b64":"fn0O2IktepZ6/4yZO+Vrjwglxo6STh8XWZNdweJnRAEW2hTU1Z1cFCq4k6T0/UBgdtzBSvzI5nzd7wWts0cZDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d9e7b4ece0138586ad7c1b14d05668a3cac068eec6870f7ce7106bb4a4a010a9","last_reissued_at":"2026-05-20T00:04:16.614164Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:04:16.614164Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Stochastic Regret Guarantees for Online Zeroth- and First-Order Bilevel Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.NA","math.NA","math.OC","math.ST","stat.TH"],"primary_cat":"cs.LG","authors_text":"Bojian Hou, Davoud Ataee Tarzanagh, George Michailidis, Li Shen, Parvin Nazari","submitted_at":"2025-11-03T00:29:36Z","abstract_excerpt":"Online bilevel optimization (OBO) is a powerful framework for machine learning problems where both outer and inner objectives evolve over time, requiring dynamic updates. Current OBO approaches rely on deterministic \\textit{window-smoothed} regret minimization, which may not accurately reflect system performance when functions change rapidly. In this work, we introduce a novel search direction and show that both first- and zeroth-order (ZO) stochastic OBO algorithms leveraging this direction achieve sublinear {stochastic bilevel regret without window smoothing}. Beyond these guarantees, our fr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.01126","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/2511.01126/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":"2511.01126","created_at":"2026-05-20T00:04:16.614302+00:00"},{"alias_kind":"arxiv_version","alias_value":"2511.01126v2","created_at":"2026-05-20T00:04:16.614302+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.01126","created_at":"2026-05-20T00:04:16.614302+00:00"},{"alias_kind":"pith_short_12","alias_value":"3HT3J3HACOCY","created_at":"2026-05-20T00:04:16.614302+00:00"},{"alias_kind":"pith_short_16","alias_value":"3HT3J3HACOCYNLL4","created_at":"2026-05-20T00:04:16.614302+00:00"},{"alias_kind":"pith_short_8","alias_value":"3HT3J3HA","created_at":"2026-05-20T00:04:16.614302+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.12693","citing_title":"IGT-OMD: Implicit Gradient Transport for Decision-Focused Learning under Delayed Feedback","ref_index":23,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3HT3J3HACOCYNLL4DMKNAVTIUP","json":"https://pith.science/pith/3HT3J3HACOCYNLL4DMKNAVTIUP.json","graph_json":"https://pith.science/api/pith-number/3HT3J3HACOCYNLL4DMKNAVTIUP/graph.json","events_json":"https://pith.science/api/pith-number/3HT3J3HACOCYNLL4DMKNAVTIUP/events.json","paper":"https://pith.science/paper/3HT3J3HA"},"agent_actions":{"view_html":"https://pith.science/pith/3HT3J3HACOCYNLL4DMKNAVTIUP","download_json":"https://pith.science/pith/3HT3J3HACOCYNLL4DMKNAVTIUP.json","view_paper":"https://pith.science/paper/3HT3J3HA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2511.01126&json=true","fetch_graph":"https://pith.science/api/pith-number/3HT3J3HACOCYNLL4DMKNAVTIUP/graph.json","fetch_events":"https://pith.science/api/pith-number/3HT3J3HACOCYNLL4DMKNAVTIUP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3HT3J3HACOCYNLL4DMKNAVTIUP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3HT3J3HACOCYNLL4DMKNAVTIUP/action/storage_attestation","attest_author":"https://pith.science/pith/3HT3J3HACOCYNLL4DMKNAVTIUP/action/author_attestation","sign_citation":"https://pith.science/pith/3HT3J3HACOCYNLL4DMKNAVTIUP/action/citation_signature","submit_replication":"https://pith.science/pith/3HT3J3HACOCYNLL4DMKNAVTIUP/action/replication_record"}},"created_at":"2026-05-20T00:04:16.614302+00:00","updated_at":"2026-05-20T00:04:16.614302+00:00"}