{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:5JRYYR6WOTLL4VWXU6GGW5F4DE","short_pith_number":"pith:5JRYYR6W","schema_version":"1.0","canonical_sha256":"ea638c47d674d6be56d7a78c6b74bc1931fd044cf0f7628392d899564ef82d53","source":{"kind":"arxiv","id":"2305.12568","version":2},"attestation_state":"computed","paper":{"title":"Det-CGD: Compressed Gradient Descent with Matrix Stepsizes for Non-Convex Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"math.OC","authors_text":"Avetik Karagulyan, Hanmin Li, Peter Richt\\'arik","submitted_at":"2023-05-21T21:09:53Z","abstract_excerpt":"This paper introduces a new method for minimizing matrix-smooth non-convex objectives through the use of novel Compressed Gradient Descent (CGD) algorithms enhanced with a matrix-valued stepsize. The proposed algorithms are theoretically analyzed first in the single-node and subsequently in the distributed settings. Our theoretical results reveal that the matrix stepsize in CGD can capture the objective's structure and lead to faster convergence compared to a scalar stepsize. As a byproduct of our general results, we emphasize the importance of selecting the compression mechanism and the matri"},"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":"2305.12568","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"math.OC","submitted_at":"2023-05-21T21:09:53Z","cross_cats_sorted":[],"title_canon_sha256":"12e5f64c8a160dd225918099615dd24c5f4ca70954ccdb9ce143616cecaa0496","abstract_canon_sha256":"6520641ed77e426bf21020a24a5493c49488a9bf9b308b345d104b54b310d832"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:10:13.921001Z","signature_b64":"Qkfqo9NVzrG0awHL+6Rhg587tvpnYORidqlmCN3dT9mwUy+hwtO1KVO+8SJpZC+IF+yOLZmMIvm6Ak9GCkilBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ea638c47d674d6be56d7a78c6b74bc1931fd044cf0f7628392d899564ef82d53","last_reissued_at":"2026-07-05T08:10:13.920572Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:10:13.920572Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Det-CGD: Compressed Gradient Descent with Matrix Stepsizes for Non-Convex Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"math.OC","authors_text":"Avetik Karagulyan, Hanmin Li, Peter Richt\\'arik","submitted_at":"2023-05-21T21:09:53Z","abstract_excerpt":"This paper introduces a new method for minimizing matrix-smooth non-convex objectives through the use of novel Compressed Gradient Descent (CGD) algorithms enhanced with a matrix-valued stepsize. The proposed algorithms are theoretically analyzed first in the single-node and subsequently in the distributed settings. Our theoretical results reveal that the matrix stepsize in CGD can capture the objective's structure and lead to faster convergence compared to a scalar stepsize. As a byproduct of our general results, we emphasize the importance of selecting the compression mechanism and the matri"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.12568","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/2305.12568/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":"2305.12568","created_at":"2026-07-05T08:10:13.920634+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.12568v2","created_at":"2026-07-05T08:10:13.920634+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.12568","created_at":"2026-07-05T08:10:13.920634+00:00"},{"alias_kind":"pith_short_12","alias_value":"5JRYYR6WOTLL","created_at":"2026-07-05T08:10:13.920634+00:00"},{"alias_kind":"pith_short_16","alias_value":"5JRYYR6WOTLL4VWX","created_at":"2026-07-05T08:10:13.920634+00:00"},{"alias_kind":"pith_short_8","alias_value":"5JRYYR6W","created_at":"2026-07-05T08:10:13.920634+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/5JRYYR6WOTLL4VWXU6GGW5F4DE","json":"https://pith.science/pith/5JRYYR6WOTLL4VWXU6GGW5F4DE.json","graph_json":"https://pith.science/api/pith-number/5JRYYR6WOTLL4VWXU6GGW5F4DE/graph.json","events_json":"https://pith.science/api/pith-number/5JRYYR6WOTLL4VWXU6GGW5F4DE/events.json","paper":"https://pith.science/paper/5JRYYR6W"},"agent_actions":{"view_html":"https://pith.science/pith/5JRYYR6WOTLL4VWXU6GGW5F4DE","download_json":"https://pith.science/pith/5JRYYR6WOTLL4VWXU6GGW5F4DE.json","view_paper":"https://pith.science/paper/5JRYYR6W","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.12568&json=true","fetch_graph":"https://pith.science/api/pith-number/5JRYYR6WOTLL4VWXU6GGW5F4DE/graph.json","fetch_events":"https://pith.science/api/pith-number/5JRYYR6WOTLL4VWXU6GGW5F4DE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5JRYYR6WOTLL4VWXU6GGW5F4DE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5JRYYR6WOTLL4VWXU6GGW5F4DE/action/storage_attestation","attest_author":"https://pith.science/pith/5JRYYR6WOTLL4VWXU6GGW5F4DE/action/author_attestation","sign_citation":"https://pith.science/pith/5JRYYR6WOTLL4VWXU6GGW5F4DE/action/citation_signature","submit_replication":"https://pith.science/pith/5JRYYR6WOTLL4VWXU6GGW5F4DE/action/replication_record"}},"created_at":"2026-07-05T08:10:13.920634+00:00","updated_at":"2026-07-05T08:10:13.920634+00:00"}