{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:KOGI6QXENOCIK66STWZEYKU6EJ","short_pith_number":"pith:KOGI6QXE","schema_version":"1.0","canonical_sha256":"538c8f42e46b84857bd29db24c2a9e2253e4fdcaed61b6468e47d89be9082736","source":{"kind":"arxiv","id":"2606.10110","version":1},"attestation_state":"computed","paper":{"title":"The Dual Averaging Power-Prox Method with Application to Heavy-Tail Incremental Gradient","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"math.OC","authors_text":"Jeremy Rack, Sebastian U. Stich, Yuan Gao","submitted_at":"2026-06-08T19:39:15Z","abstract_excerpt":"We study finite-sum composite optimization under two departures from classical stochastic gradient descent theory that are central in practice: incremental gradient access and heavy-tailed gradient noise. Specifically, we consider fixed cyclic passes over component gradients and assume that, at the optimum, component gradients have a bounded $q$-th centralized moment for some $q\\in(1,2]$. This setting is much closer to modern ML training practice than the assumptions used in classical SGD theory, yet its theoretical understanding remains limited. We propose a Dual Averaging Power-Prox method f"},"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":"2606.10110","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"math.OC","submitted_at":"2026-06-08T19:39:15Z","cross_cats_sorted":[],"title_canon_sha256":"2c5f21ad5ebb51e2cee8ec0dd84f56ab956c5b50167e6dcd9c7ba3b7504e3b53","abstract_canon_sha256":"1bac540ef18027ca642761a497e3af0134ce20a4d2ed5d5d1ee5125ded014bfd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-10T01:08:54.748306Z","signature_b64":"HRW/JG3UzS1+/RxLryFCEqR0OOAVzPacs1M3zJA0ZzHa2iU//zoySDI46W6d9fm1X5DbselR0jQ7yK0HBbnfCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"538c8f42e46b84857bd29db24c2a9e2253e4fdcaed61b6468e47d89be9082736","last_reissued_at":"2026-06-10T01:08:54.747392Z","signature_status":"signed_v1","first_computed_at":"2026-06-10T01:08:54.747392Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Dual Averaging Power-Prox Method with Application to Heavy-Tail Incremental Gradient","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"math.OC","authors_text":"Jeremy Rack, Sebastian U. Stich, Yuan Gao","submitted_at":"2026-06-08T19:39:15Z","abstract_excerpt":"We study finite-sum composite optimization under two departures from classical stochastic gradient descent theory that are central in practice: incremental gradient access and heavy-tailed gradient noise. Specifically, we consider fixed cyclic passes over component gradients and assume that, at the optimum, component gradients have a bounded $q$-th centralized moment for some $q\\in(1,2]$. This setting is much closer to modern ML training practice than the assumptions used in classical SGD theory, yet its theoretical understanding remains limited. We propose a Dual Averaging Power-Prox method f"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.10110","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/2606.10110/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":"2606.10110","created_at":"2026-06-10T01:08:54.747522+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.10110v1","created_at":"2026-06-10T01:08:54.747522+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.10110","created_at":"2026-06-10T01:08:54.747522+00:00"},{"alias_kind":"pith_short_12","alias_value":"KOGI6QXENOCI","created_at":"2026-06-10T01:08:54.747522+00:00"},{"alias_kind":"pith_short_16","alias_value":"KOGI6QXENOCIK66S","created_at":"2026-06-10T01:08:54.747522+00:00"},{"alias_kind":"pith_short_8","alias_value":"KOGI6QXE","created_at":"2026-06-10T01:08:54.747522+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/KOGI6QXENOCIK66STWZEYKU6EJ","json":"https://pith.science/pith/KOGI6QXENOCIK66STWZEYKU6EJ.json","graph_json":"https://pith.science/api/pith-number/KOGI6QXENOCIK66STWZEYKU6EJ/graph.json","events_json":"https://pith.science/api/pith-number/KOGI6QXENOCIK66STWZEYKU6EJ/events.json","paper":"https://pith.science/paper/KOGI6QXE"},"agent_actions":{"view_html":"https://pith.science/pith/KOGI6QXENOCIK66STWZEYKU6EJ","download_json":"https://pith.science/pith/KOGI6QXENOCIK66STWZEYKU6EJ.json","view_paper":"https://pith.science/paper/KOGI6QXE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.10110&json=true","fetch_graph":"https://pith.science/api/pith-number/KOGI6QXENOCIK66STWZEYKU6EJ/graph.json","fetch_events":"https://pith.science/api/pith-number/KOGI6QXENOCIK66STWZEYKU6EJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KOGI6QXENOCIK66STWZEYKU6EJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KOGI6QXENOCIK66STWZEYKU6EJ/action/storage_attestation","attest_author":"https://pith.science/pith/KOGI6QXENOCIK66STWZEYKU6EJ/action/author_attestation","sign_citation":"https://pith.science/pith/KOGI6QXENOCIK66STWZEYKU6EJ/action/citation_signature","submit_replication":"https://pith.science/pith/KOGI6QXENOCIK66STWZEYKU6EJ/action/replication_record"}},"created_at":"2026-06-10T01:08:54.747522+00:00","updated_at":"2026-06-10T01:08:54.747522+00:00"}