{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:MQCWZPQ5E4272YIV4CF2KJ7YYU","short_pith_number":"pith:MQCWZPQ5","schema_version":"1.0","canonical_sha256":"64056cbe1d2735fd6115e08ba527f8c510fd8140deb58b73944398c30b1f2f91","source":{"kind":"arxiv","id":"2208.11195","version":1},"attestation_state":"computed","paper":{"title":"Robustness to Unbounded Smoothness of Generalized SignSGD","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.LG","authors_text":"Francesco Orabona, Michael Crawshaw, Mingrui Liu, Wei Zhang, Zhenxun Zhuang","submitted_at":"2022-08-23T21:11:19Z","abstract_excerpt":"Traditional analyses in non-convex optimization typically rely on the smoothness assumption, namely requiring the gradients to be Lipschitz. However, recent evidence shows that this smoothness condition does not capture the properties of some deep learning objective functions, including the ones involving Recurrent Neural Networks and LSTMs. Instead, they satisfy a much more relaxed condition, with potentially unbounded smoothness. Under this relaxed assumption, it has been theoretically and empirically shown that the gradient-clipped SGD has an advantage over the vanilla one. In this paper, w"},"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":"2208.11195","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-08-23T21:11:19Z","cross_cats_sorted":["math.OC"],"title_canon_sha256":"5f3dc95485ef126b0a32fec24eae329e875030d886d8c139c0bcf36b148f6bf1","abstract_canon_sha256":"bbde6ea1c5fd65af4a1dbf052f06d34556544b35dacbccffd67da1d6fe2e8efe"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:51:12.388469Z","signature_b64":"gOW48olwVYszxowI2A369gjwQdsa5T4Z/I6YzVg0AKZbkHe5Rty38KNnj+ufhprdYezYhVK/4jpr3PyWpZzIDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"64056cbe1d2735fd6115e08ba527f8c510fd8140deb58b73944398c30b1f2f91","last_reissued_at":"2026-07-05T04:51:12.387945Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:51:12.387945Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Robustness to Unbounded Smoothness of Generalized SignSGD","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.LG","authors_text":"Francesco Orabona, Michael Crawshaw, Mingrui Liu, Wei Zhang, Zhenxun Zhuang","submitted_at":"2022-08-23T21:11:19Z","abstract_excerpt":"Traditional analyses in non-convex optimization typically rely on the smoothness assumption, namely requiring the gradients to be Lipschitz. However, recent evidence shows that this smoothness condition does not capture the properties of some deep learning objective functions, including the ones involving Recurrent Neural Networks and LSTMs. Instead, they satisfy a much more relaxed condition, with potentially unbounded smoothness. Under this relaxed assumption, it has been theoretically and empirically shown that the gradient-clipped SGD has an advantage over the vanilla one. In this paper, w"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2208.11195","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/2208.11195/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":"2208.11195","created_at":"2026-07-05T04:51:12.388009+00:00"},{"alias_kind":"arxiv_version","alias_value":"2208.11195v1","created_at":"2026-07-05T04:51:12.388009+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2208.11195","created_at":"2026-07-05T04:51:12.388009+00:00"},{"alias_kind":"pith_short_12","alias_value":"MQCWZPQ5E427","created_at":"2026-07-05T04:51:12.388009+00:00"},{"alias_kind":"pith_short_16","alias_value":"MQCWZPQ5E4272YIV","created_at":"2026-07-05T04:51:12.388009+00:00"},{"alias_kind":"pith_short_8","alias_value":"MQCWZPQ5","created_at":"2026-07-05T04:51:12.388009+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/MQCWZPQ5E4272YIV4CF2KJ7YYU","json":"https://pith.science/pith/MQCWZPQ5E4272YIV4CF2KJ7YYU.json","graph_json":"https://pith.science/api/pith-number/MQCWZPQ5E4272YIV4CF2KJ7YYU/graph.json","events_json":"https://pith.science/api/pith-number/MQCWZPQ5E4272YIV4CF2KJ7YYU/events.json","paper":"https://pith.science/paper/MQCWZPQ5"},"agent_actions":{"view_html":"https://pith.science/pith/MQCWZPQ5E4272YIV4CF2KJ7YYU","download_json":"https://pith.science/pith/MQCWZPQ5E4272YIV4CF2KJ7YYU.json","view_paper":"https://pith.science/paper/MQCWZPQ5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2208.11195&json=true","fetch_graph":"https://pith.science/api/pith-number/MQCWZPQ5E4272YIV4CF2KJ7YYU/graph.json","fetch_events":"https://pith.science/api/pith-number/MQCWZPQ5E4272YIV4CF2KJ7YYU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MQCWZPQ5E4272YIV4CF2KJ7YYU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MQCWZPQ5E4272YIV4CF2KJ7YYU/action/storage_attestation","attest_author":"https://pith.science/pith/MQCWZPQ5E4272YIV4CF2KJ7YYU/action/author_attestation","sign_citation":"https://pith.science/pith/MQCWZPQ5E4272YIV4CF2KJ7YYU/action/citation_signature","submit_replication":"https://pith.science/pith/MQCWZPQ5E4272YIV4CF2KJ7YYU/action/replication_record"}},"created_at":"2026-07-05T04:51:12.388009+00:00","updated_at":"2026-07-05T04:51:12.388009+00:00"}