{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:EQALOI3D7APAT7YKMUO4EXFJX5","short_pith_number":"pith:EQALOI3D","schema_version":"1.0","canonical_sha256":"2400b72363f81e09ff0a651dc25ca9bf4a5247c64cc45de062a5152016852e7f","source":{"kind":"arxiv","id":"2606.21354","version":1},"attestation_state":"computed","paper":{"title":"Restart and Adaptive Acceleration in Stochastic Gradient Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"math.OC","authors_text":"Alexandre d'Aspremont, Ali Elhishi, Chistophe Roux","submitted_at":"2026-06-19T11:54:49Z","abstract_excerpt":"We study restart schemes in stochastic optimization problems for non-smooth and weakly convex that satisfy a Kurdyka-\\L ojasiewicz inequality. We show that using restarts allows us to leverage the K{\\L} inequalities to achieve improved rates of convergence, with acceleration depending explicitly on the K{\\L} exponent. Furthermore, optimal restart schedules lead to learning-rates akin to Polyak steps for SGD. While regularity constants such as the K{\\L} exponent are typically unknown in practice, we prove that restart schemes are robust to a significant misspecification of these constants, henc"},"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":"2606.21354","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"math.OC","submitted_at":"2026-06-19T11:54:49Z","cross_cats_sorted":[],"title_canon_sha256":"84502c67915bc787959aa68391588b143182ff2660ffa29cf6f1eb134a139dc7","abstract_canon_sha256":"ddee77a2f88f63f5c571019d9e6362879757b3921ea32f938269e807a9438c47"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T01:12:38.261301Z","signature_b64":"mJQn+WQh+l+CC0q6rDc1WC0uorRveqUfAiYDObETAve3zwAtO01QXmipE4MXgs+B1iyPbw8RCkbtgSkBYjNbAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2400b72363f81e09ff0a651dc25ca9bf4a5247c64cc45de062a5152016852e7f","last_reissued_at":"2026-06-23T01:12:38.260793Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T01:12:38.260793Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Restart and Adaptive Acceleration in Stochastic Gradient Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"math.OC","authors_text":"Alexandre d'Aspremont, Ali Elhishi, Chistophe Roux","submitted_at":"2026-06-19T11:54:49Z","abstract_excerpt":"We study restart schemes in stochastic optimization problems for non-smooth and weakly convex that satisfy a Kurdyka-\\L ojasiewicz inequality. We show that using restarts allows us to leverage the K{\\L} inequalities to achieve improved rates of convergence, with acceleration depending explicitly on the K{\\L} exponent. Furthermore, optimal restart schedules lead to learning-rates akin to Polyak steps for SGD. While regularity constants such as the K{\\L} exponent are typically unknown in practice, we prove that restart schemes are robust to a significant misspecification of these constants, henc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.21354","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/2606.21354/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":"2606.21354","created_at":"2026-06-23T01:12:38.260870+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.21354v1","created_at":"2026-06-23T01:12:38.260870+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.21354","created_at":"2026-06-23T01:12:38.260870+00:00"},{"alias_kind":"pith_short_12","alias_value":"EQALOI3D7APA","created_at":"2026-06-23T01:12:38.260870+00:00"},{"alias_kind":"pith_short_16","alias_value":"EQALOI3D7APAT7YK","created_at":"2026-06-23T01:12:38.260870+00:00"},{"alias_kind":"pith_short_8","alias_value":"EQALOI3D","created_at":"2026-06-23T01:12:38.260870+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/EQALOI3D7APAT7YKMUO4EXFJX5","json":"https://pith.science/pith/EQALOI3D7APAT7YKMUO4EXFJX5.json","graph_json":"https://pith.science/api/pith-number/EQALOI3D7APAT7YKMUO4EXFJX5/graph.json","events_json":"https://pith.science/api/pith-number/EQALOI3D7APAT7YKMUO4EXFJX5/events.json","paper":"https://pith.science/paper/EQALOI3D"},"agent_actions":{"view_html":"https://pith.science/pith/EQALOI3D7APAT7YKMUO4EXFJX5","download_json":"https://pith.science/pith/EQALOI3D7APAT7YKMUO4EXFJX5.json","view_paper":"https://pith.science/paper/EQALOI3D","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.21354&json=true","fetch_graph":"https://pith.science/api/pith-number/EQALOI3D7APAT7YKMUO4EXFJX5/graph.json","fetch_events":"https://pith.science/api/pith-number/EQALOI3D7APAT7YKMUO4EXFJX5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EQALOI3D7APAT7YKMUO4EXFJX5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EQALOI3D7APAT7YKMUO4EXFJX5/action/storage_attestation","attest_author":"https://pith.science/pith/EQALOI3D7APAT7YKMUO4EXFJX5/action/author_attestation","sign_citation":"https://pith.science/pith/EQALOI3D7APAT7YKMUO4EXFJX5/action/citation_signature","submit_replication":"https://pith.science/pith/EQALOI3D7APAT7YKMUO4EXFJX5/action/replication_record"}},"created_at":"2026-06-23T01:12:38.260870+00:00","updated_at":"2026-06-23T01:12:38.260870+00:00"}