{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6Q7Y555O5H6GQKC4FSACK3FGQW","short_pith_number":"pith:6Q7Y555O","schema_version":"1.0","canonical_sha256":"f43f8ef7aee9fc68285c2c80256ca68599afead52572929a61d7b68b4c7d8942","source":{"kind":"arxiv","id":"2405.17834","version":3},"attestation_state":"computed","paper":{"title":"Revisiting Step-Size Assumptions in Stochastic Approximation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML","stat.TH"],"primary_cat":"math.ST","authors_text":"Caio Kalil Lauand, Sean Meyn","submitted_at":"2024-05-28T05:11:05Z","abstract_excerpt":"Many machine learning and optimization algorithms are built upon the framework of stochastic approximation (SA), for which the selection of step-size (or learning rate) $\\{\\alpha_n\\}$ is crucial for success. An essential condition for convergence is the assumption that $\\sum_n \\alpha_n = \\infty$. Moreover, in all theory to date it is assumed that $\\sum_n \\alpha_n^2 < \\infty$ (the sequence is square summable). In this paper it is shown for the first time that this assumption is not required for convergence and finer results.\n  The main results are restricted to the special case $\\alpha_n = \\alp"},"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":"2405.17834","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"math.ST","submitted_at":"2024-05-28T05:11:05Z","cross_cats_sorted":["stat.ML","stat.TH"],"title_canon_sha256":"5220b90fbec00ab744ff443dcfb0cb0e2e9f88739dc3222aa17e9543aeab25a2","abstract_canon_sha256":"8b91d7332b9c0e3b5d4e26ac7e68c52609c472fbf6f54e37394aca71986cc8bc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:47:11.836084Z","signature_b64":"UlNJbbOZLG6lNSLCeIbIifG97RvcEnWNkhhlzI2CyParJ0t8S8N5hXMb6gQACh4XdJWv22P813ZAgOOvIqgBAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f43f8ef7aee9fc68285c2c80256ca68599afead52572929a61d7b68b4c7d8942","last_reissued_at":"2026-07-05T11:47:11.835554Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:47:11.835554Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Revisiting Step-Size Assumptions in Stochastic Approximation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML","stat.TH"],"primary_cat":"math.ST","authors_text":"Caio Kalil Lauand, Sean Meyn","submitted_at":"2024-05-28T05:11:05Z","abstract_excerpt":"Many machine learning and optimization algorithms are built upon the framework of stochastic approximation (SA), for which the selection of step-size (or learning rate) $\\{\\alpha_n\\}$ is crucial for success. An essential condition for convergence is the assumption that $\\sum_n \\alpha_n = \\infty$. Moreover, in all theory to date it is assumed that $\\sum_n \\alpha_n^2 < \\infty$ (the sequence is square summable). In this paper it is shown for the first time that this assumption is not required for convergence and finer results.\n  The main results are restricted to the special case $\\alpha_n = \\alp"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.17834","kind":"arxiv","version":3},"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/2405.17834/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":"2405.17834","created_at":"2026-07-05T11:47:11.835629+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.17834v3","created_at":"2026-07-05T11:47:11.835629+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.17834","created_at":"2026-07-05T11:47:11.835629+00:00"},{"alias_kind":"pith_short_12","alias_value":"6Q7Y555O5H6G","created_at":"2026-07-05T11:47:11.835629+00:00"},{"alias_kind":"pith_short_16","alias_value":"6Q7Y555O5H6GQKC4","created_at":"2026-07-05T11:47:11.835629+00:00"},{"alias_kind":"pith_short_8","alias_value":"6Q7Y555O","created_at":"2026-07-05T11:47:11.835629+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.31309","citing_title":"Non-Asymptotic Convergence of Stochastic Iterative Algorithms: A Lyapunov Framework","ref_index":70,"is_internal_anchor":false},{"citing_arxiv_id":"2504.18743","citing_title":"From Set Convergence to Pointwise Convergence: Finite-Time Guarantees for Average-Reward Q-Learning with Adaptive Stepsizes","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13378","citing_title":"Revisiting the Constant Stepsize Stochastic Approximation with Decision-Dependent Markovian Noise","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6Q7Y555O5H6GQKC4FSACK3FGQW","json":"https://pith.science/pith/6Q7Y555O5H6GQKC4FSACK3FGQW.json","graph_json":"https://pith.science/api/pith-number/6Q7Y555O5H6GQKC4FSACK3FGQW/graph.json","events_json":"https://pith.science/api/pith-number/6Q7Y555O5H6GQKC4FSACK3FGQW/events.json","paper":"https://pith.science/paper/6Q7Y555O"},"agent_actions":{"view_html":"https://pith.science/pith/6Q7Y555O5H6GQKC4FSACK3FGQW","download_json":"https://pith.science/pith/6Q7Y555O5H6GQKC4FSACK3FGQW.json","view_paper":"https://pith.science/paper/6Q7Y555O","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.17834&json=true","fetch_graph":"https://pith.science/api/pith-number/6Q7Y555O5H6GQKC4FSACK3FGQW/graph.json","fetch_events":"https://pith.science/api/pith-number/6Q7Y555O5H6GQKC4FSACK3FGQW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6Q7Y555O5H6GQKC4FSACK3FGQW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6Q7Y555O5H6GQKC4FSACK3FGQW/action/storage_attestation","attest_author":"https://pith.science/pith/6Q7Y555O5H6GQKC4FSACK3FGQW/action/author_attestation","sign_citation":"https://pith.science/pith/6Q7Y555O5H6GQKC4FSACK3FGQW/action/citation_signature","submit_replication":"https://pith.science/pith/6Q7Y555O5H6GQKC4FSACK3FGQW/action/replication_record"}},"created_at":"2026-07-05T11:47:11.835629+00:00","updated_at":"2026-07-05T11:47:11.835629+00:00"}