{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:PVWGRL77U7E3HYQWHS3UXAVKSJ","short_pith_number":"pith:PVWGRL77","schema_version":"1.0","canonical_sha256":"7d6c68afffa7c9b3e2163cb74b82aa927370116ede88e7f3dbcb873386758a10","source":{"kind":"arxiv","id":"2605.19781","version":1},"attestation_state":"computed","paper":{"title":"From SGD to Muon: Adaptive Optimization via Schatten-p Norms","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Corentin Friedrich, DTIPG - SNCF, Mathieu Serrurier (IRIT), Thomas Massena (IRIT, UT3)","submitted_at":"2026-05-19T12:47:41Z","abstract_excerpt":"Modern optimizers, like Muon, impose matrix-wise geometry constraints on their updates. These matrix-wise constraints can be unified under Linear Minimization Oracle (LMO) theory. However, all current methods impose fixed LMO geometries for the update rules, chosen by-design or empirically, which are not necessarily optimal according to the problem's geometry. We introduce a novel efficient datadriven criterion for dynamically choosing proxy-optimal update LMO geometries on individual Deep Neural Network layers. Derived in closed form from gradient and activation statistics using a single-step"},"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":"2605.19781","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-19T12:47:41Z","cross_cats_sorted":[],"title_canon_sha256":"7f0968ba2fa16eb1160d40316134b8541807ee5a0e5dab455436be81cc914a9a","abstract_canon_sha256":"93753c1c6cfa1077b21bcf0070a736df21b266f218cb190cbe124ad4d1f611a7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:06:13.715914Z","signature_b64":"7lomlqFWYx1tOSWpJ1hk1kEn9nyfd9eUK5N8QxQECcf2qLgPFDQ7DAdqOCdQ2BSFvKWaq6U9XxhRBdEzL9YBAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7d6c68afffa7c9b3e2163cb74b82aa927370116ede88e7f3dbcb873386758a10","last_reissued_at":"2026-05-20T01:06:13.715063Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:06:13.715063Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From SGD to Muon: Adaptive Optimization via Schatten-p Norms","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Corentin Friedrich, DTIPG - SNCF, Mathieu Serrurier (IRIT), Thomas Massena (IRIT, UT3)","submitted_at":"2026-05-19T12:47:41Z","abstract_excerpt":"Modern optimizers, like Muon, impose matrix-wise geometry constraints on their updates. These matrix-wise constraints can be unified under Linear Minimization Oracle (LMO) theory. However, all current methods impose fixed LMO geometries for the update rules, chosen by-design or empirically, which are not necessarily optimal according to the problem's geometry. We introduce a novel efficient datadriven criterion for dynamically choosing proxy-optimal update LMO geometries on individual Deep Neural Network layers. Derived in closed form from gradient and activation statistics using a single-step"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.19781","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/2605.19781/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":"2605.19781","created_at":"2026-05-20T01:06:13.715210+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.19781v1","created_at":"2026-05-20T01:06:13.715210+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.19781","created_at":"2026-05-20T01:06:13.715210+00:00"},{"alias_kind":"pith_short_12","alias_value":"PVWGRL77U7E3","created_at":"2026-05-20T01:06:13.715210+00:00"},{"alias_kind":"pith_short_16","alias_value":"PVWGRL77U7E3HYQW","created_at":"2026-05-20T01:06:13.715210+00:00"},{"alias_kind":"pith_short_8","alias_value":"PVWGRL77","created_at":"2026-05-20T01:06:13.715210+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/PVWGRL77U7E3HYQWHS3UXAVKSJ","json":"https://pith.science/pith/PVWGRL77U7E3HYQWHS3UXAVKSJ.json","graph_json":"https://pith.science/api/pith-number/PVWGRL77U7E3HYQWHS3UXAVKSJ/graph.json","events_json":"https://pith.science/api/pith-number/PVWGRL77U7E3HYQWHS3UXAVKSJ/events.json","paper":"https://pith.science/paper/PVWGRL77"},"agent_actions":{"view_html":"https://pith.science/pith/PVWGRL77U7E3HYQWHS3UXAVKSJ","download_json":"https://pith.science/pith/PVWGRL77U7E3HYQWHS3UXAVKSJ.json","view_paper":"https://pith.science/paper/PVWGRL77","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.19781&json=true","fetch_graph":"https://pith.science/api/pith-number/PVWGRL77U7E3HYQWHS3UXAVKSJ/graph.json","fetch_events":"https://pith.science/api/pith-number/PVWGRL77U7E3HYQWHS3UXAVKSJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PVWGRL77U7E3HYQWHS3UXAVKSJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PVWGRL77U7E3HYQWHS3UXAVKSJ/action/storage_attestation","attest_author":"https://pith.science/pith/PVWGRL77U7E3HYQWHS3UXAVKSJ/action/author_attestation","sign_citation":"https://pith.science/pith/PVWGRL77U7E3HYQWHS3UXAVKSJ/action/citation_signature","submit_replication":"https://pith.science/pith/PVWGRL77U7E3HYQWHS3UXAVKSJ/action/replication_record"}},"created_at":"2026-05-20T01:06:13.715210+00:00","updated_at":"2026-05-20T01:06:13.715210+00:00"}