{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:6AYKRCMXVUEZOSOVLFDHYBL3DU","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"31402d2ae3ca535d26562b263de46ed4211d596da4d05f7cda2dc4d783f213d9","cross_cats_sorted":["math.OC","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-03-16T20:49:34Z","title_canon_sha256":"02f556139f0ea6494085dee48aed5eb68d8a163f387d77b3ce8c83eccc61c55f"},"schema_version":"1.0","source":{"id":"2503.12645","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.12645","created_at":"2026-07-05T10:46:02Z"},{"alias_kind":"arxiv_version","alias_value":"2503.12645v2","created_at":"2026-07-05T10:46:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.12645","created_at":"2026-07-05T10:46:02Z"},{"alias_kind":"pith_short_12","alias_value":"6AYKRCMXVUEZ","created_at":"2026-07-05T10:46:02Z"},{"alias_kind":"pith_short_16","alias_value":"6AYKRCMXVUEZOSOV","created_at":"2026-07-05T10:46:02Z"},{"alias_kind":"pith_short_8","alias_value":"6AYKRCMX","created_at":"2026-07-05T10:46:02Z"}],"graph_snapshots":[{"event_id":"sha256:1b68bd3b902de08a691fda868183703b231cb69e7c48e51454442c930e628b64","target":"graph","created_at":"2026-07-05T10:46:02Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2503.12645/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Optimization with matrix gradient orthogonalization has recently demonstrated impressive results in the training of deep neural networks (Jordan et al., 2024; Liu et al., 2025). In this paper, we provide a theoretical analysis of this approach. In particular, we show that the orthogonalized gradient method can be seen as a first-order trust-region optimization method, where the trust-region is defined in terms of the matrix spectral norm. Motivated by this observation, we develop the stochastic non-Euclidean trust-region gradient method with momentum, which recovers the Muon optimizer (Jordan ","authors_text":"Dmitry Kovalev","cross_cats":["math.OC","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-03-16T20:49:34Z","title":"Understanding Gradient Orthogonalization for Deep Learning via Non-Euclidean Trust-Region Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.12645","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:24671ef19d43145d3de878b42911065e02f44a9d4d74b3c633c3235ff716f5b8","target":"record","created_at":"2026-07-05T10:46:02Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"31402d2ae3ca535d26562b263de46ed4211d596da4d05f7cda2dc4d783f213d9","cross_cats_sorted":["math.OC","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-03-16T20:49:34Z","title_canon_sha256":"02f556139f0ea6494085dee48aed5eb68d8a163f387d77b3ce8c83eccc61c55f"},"schema_version":"1.0","source":{"id":"2503.12645","kind":"arxiv","version":2}},"canonical_sha256":"f030a88997ad099749d559467c057b1d0dedaf5811a02693d0305a2d45fbccdd","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f030a88997ad099749d559467c057b1d0dedaf5811a02693d0305a2d45fbccdd","first_computed_at":"2026-07-05T10:46:02.376500Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:46:02.376500Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8ySP3u1h7d42c7A2pizo8mIPZBNqtoFo/nr4zcPRlN4VtEvaN2nxWRU0Vvrf4wXdJLcBI2doBrXZML2nNGzsDA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:46:02.376960Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.12645","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:24671ef19d43145d3de878b42911065e02f44a9d4d74b3c633c3235ff716f5b8","sha256:1b68bd3b902de08a691fda868183703b231cb69e7c48e51454442c930e628b64"],"state_sha256":"a357beafb71e94a06e763680b08067218cc83b0bbf145f4f4b6e4e393a3a3f4e"}