{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:WSBWPGLTFP7RRRZYE5QDQJ7GOR","short_pith_number":"pith:WSBWPGLT","schema_version":"1.0","canonical_sha256":"b4836799732bff18c73827603827e6744b441338e77916b04621b5bf5c28908e","source":{"kind":"arxiv","id":"2502.07923","version":2},"attestation_state":"computed","paper":{"title":"Sign Operator for Coping with Heavy-Tailed Noise in Non-Convex Optimization: High Probability Bounds Under $(L_0, L_1)$-Smoothness","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"math.OC","authors_text":"Alexander Beznosikov, Alexander Gasnikov, Andrei Semenov, Mark Ikonnikov, Nikita Kornilov, Philip Zmushko","submitted_at":"2025-02-11T19:54:11Z","abstract_excerpt":"In recent years, non-convex optimization problems are more often described by generalized $(L_0, L_1)$-smoothness assumption rather than standard one. Meanwhile, severely corrupted data used in these problems has increased the demand for methods capable of handling heavy-tailed noises, i.e., noises with bounded $\\kappa$-th moment. Motivated by these real-world trends and challenges, we explore sign-based methods in this setup and demonstrate their effectiveness in comparison with other popular solutions like clipping or normalization.\n  In theory, we prove the first-known high probability conv"},"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":"2502.07923","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"math.OC","submitted_at":"2025-02-11T19:54:11Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"7f6942cafaa89b59cc5d401945e20b516bb012a15aa05155bfd75c6ab342d017","abstract_canon_sha256":"bc8e71a412b345a6169504acf3894dfd147d4cf5ccd7a8fb85956a234019e64a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:10:35.093608Z","signature_b64":"Qjz71YW/BBasxGGc3ax6U9Xt2YfCvyalmJsndKMRxUKn0GKMU8zRtEBo8RgDLOnHhEeC0qBMKzf5/QZdY+wECA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b4836799732bff18c73827603827e6744b441338e77916b04621b5bf5c28908e","last_reissued_at":"2026-07-05T11:10:35.092951Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:10:35.092951Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Sign Operator for Coping with Heavy-Tailed Noise in Non-Convex Optimization: High Probability Bounds Under $(L_0, L_1)$-Smoothness","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"math.OC","authors_text":"Alexander Beznosikov, Alexander Gasnikov, Andrei Semenov, Mark Ikonnikov, Nikita Kornilov, Philip Zmushko","submitted_at":"2025-02-11T19:54:11Z","abstract_excerpt":"In recent years, non-convex optimization problems are more often described by generalized $(L_0, L_1)$-smoothness assumption rather than standard one. Meanwhile, severely corrupted data used in these problems has increased the demand for methods capable of handling heavy-tailed noises, i.e., noises with bounded $\\kappa$-th moment. Motivated by these real-world trends and challenges, we explore sign-based methods in this setup and demonstrate their effectiveness in comparison with other popular solutions like clipping or normalization.\n  In theory, we prove the first-known high probability conv"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.07923","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/2502.07923/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":"2502.07923","created_at":"2026-07-05T11:10:35.093028+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.07923v2","created_at":"2026-07-05T11:10:35.093028+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.07923","created_at":"2026-07-05T11:10:35.093028+00:00"},{"alias_kind":"pith_short_12","alias_value":"WSBWPGLTFP7R","created_at":"2026-07-05T11:10:35.093028+00:00"},{"alias_kind":"pith_short_16","alias_value":"WSBWPGLTFP7RRRZY","created_at":"2026-07-05T11:10:35.093028+00:00"},{"alias_kind":"pith_short_8","alias_value":"WSBWPGLT","created_at":"2026-07-05T11:10:35.093028+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.19811","citing_title":"LionMuon: Alternating Spectral and Sign Descent for Efficient Training","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31371","citing_title":"Softsign: Smooth Sign in Your Optimizer For Better Parameter Heterogeneity Handling","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2510.06141","citing_title":"High-Probability Convergence Guarantees of Decentralized SGD","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19811","citing_title":"LionMuon: Alternating Spectral and Sign Descent for Efficient Training","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2510.06141","citing_title":"High-Probability Convergence Guarantees of Decentralized SGD","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11850","citing_title":"Constrained Stochastic Spectral Preconditioning Converges for Nonconvex Objectives","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06615","citing_title":"When and Why SignSGD Outperforms SGD: A Theoretical Study Based on $\\ell_1$-norm Lower Bounds","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WSBWPGLTFP7RRRZYE5QDQJ7GOR","json":"https://pith.science/pith/WSBWPGLTFP7RRRZYE5QDQJ7GOR.json","graph_json":"https://pith.science/api/pith-number/WSBWPGLTFP7RRRZYE5QDQJ7GOR/graph.json","events_json":"https://pith.science/api/pith-number/WSBWPGLTFP7RRRZYE5QDQJ7GOR/events.json","paper":"https://pith.science/paper/WSBWPGLT"},"agent_actions":{"view_html":"https://pith.science/pith/WSBWPGLTFP7RRRZYE5QDQJ7GOR","download_json":"https://pith.science/pith/WSBWPGLTFP7RRRZYE5QDQJ7GOR.json","view_paper":"https://pith.science/paper/WSBWPGLT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.07923&json=true","fetch_graph":"https://pith.science/api/pith-number/WSBWPGLTFP7RRRZYE5QDQJ7GOR/graph.json","fetch_events":"https://pith.science/api/pith-number/WSBWPGLTFP7RRRZYE5QDQJ7GOR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WSBWPGLTFP7RRRZYE5QDQJ7GOR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WSBWPGLTFP7RRRZYE5QDQJ7GOR/action/storage_attestation","attest_author":"https://pith.science/pith/WSBWPGLTFP7RRRZYE5QDQJ7GOR/action/author_attestation","sign_citation":"https://pith.science/pith/WSBWPGLTFP7RRRZYE5QDQJ7GOR/action/citation_signature","submit_replication":"https://pith.science/pith/WSBWPGLTFP7RRRZYE5QDQJ7GOR/action/replication_record"}},"created_at":"2026-07-05T11:10:35.093028+00:00","updated_at":"2026-07-05T11:10:35.093028+00:00"}