{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:6K5VMBJYBWRFZ7HDRWSSYTQ3P5","short_pith_number":"pith:6K5VMBJY","schema_version":"1.0","canonical_sha256":"f2bb5605380da25cfce38da52c4e1b7f6dcb309679a41a9e9e3e5f4a02e2f832","source":{"kind":"arxiv","id":"2606.12921","version":1},"attestation_state":"computed","paper":{"title":"LoRA-Muon: Spectral Steepest Descent on the Low-Rank Manifold","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"C\\'edric Simal, Franz Louis Cesista, Katherine Crowson, Stella Biderman","submitted_at":"2026-06-11T05:24:45Z","abstract_excerpt":"Low-Rank Adaptation (LoRA) significantly reduces compute and memory costs for finetuning Deep Learning models but is often harder to tune than dense training: when using factor-wise optimizers such as AdamW, it is sensitive to initialization choices, its optimal learning rates transfer poorly across ranks, and it often fails to beat dense baselines. We derive LoRA-Muon by applying the Muon optimizer's spectral steepest-descent rule to the low-rank setting. Along with our split weight-decay rule, our main claim is that LoRA-Muon is a good low-rank proxy for full-rank Muon and Shampoo-family opt"},"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":"2606.12921","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-11T05:24:45Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"9e2ca83f932ecb681449f4ba1fdd738ff9f4d3b38cac454a9694d16e104092d7","abstract_canon_sha256":"4400a614da6315007253871fd85e93bc9e132e945edf5376c6f4675e6d584f89"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-12T01:09:34.472961Z","signature_b64":"NOEQWL+GYrN31+RRrywJa8v48MZ4hHNlnxTAQXw23VDaUNv5pMxtv1tncFXOVnK+EcfIxTZSDlF5kmyOgnWXAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f2bb5605380da25cfce38da52c4e1b7f6dcb309679a41a9e9e3e5f4a02e2f832","last_reissued_at":"2026-06-12T01:09:34.472588Z","signature_status":"signed_v1","first_computed_at":"2026-06-12T01:09:34.472588Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LoRA-Muon: Spectral Steepest Descent on the Low-Rank Manifold","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"C\\'edric Simal, Franz Louis Cesista, Katherine Crowson, Stella Biderman","submitted_at":"2026-06-11T05:24:45Z","abstract_excerpt":"Low-Rank Adaptation (LoRA) significantly reduces compute and memory costs for finetuning Deep Learning models but is often harder to tune than dense training: when using factor-wise optimizers such as AdamW, it is sensitive to initialization choices, its optimal learning rates transfer poorly across ranks, and it often fails to beat dense baselines. We derive LoRA-Muon by applying the Muon optimizer's spectral steepest-descent rule to the low-rank setting. Along with our split weight-decay rule, our main claim is that LoRA-Muon is a good low-rank proxy for full-rank Muon and Shampoo-family opt"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.12921","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/2606.12921/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":"2606.12921","created_at":"2026-06-12T01:09:34.472647+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.12921v1","created_at":"2026-06-12T01:09:34.472647+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.12921","created_at":"2026-06-12T01:09:34.472647+00:00"},{"alias_kind":"pith_short_12","alias_value":"6K5VMBJYBWRF","created_at":"2026-06-12T01:09:34.472647+00:00"},{"alias_kind":"pith_short_16","alias_value":"6K5VMBJYBWRFZ7HD","created_at":"2026-06-12T01:09:34.472647+00:00"},{"alias_kind":"pith_short_8","alias_value":"6K5VMBJY","created_at":"2026-06-12T01:09:34.472647+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/6K5VMBJYBWRFZ7HDRWSSYTQ3P5","json":"https://pith.science/pith/6K5VMBJYBWRFZ7HDRWSSYTQ3P5.json","graph_json":"https://pith.science/api/pith-number/6K5VMBJYBWRFZ7HDRWSSYTQ3P5/graph.json","events_json":"https://pith.science/api/pith-number/6K5VMBJYBWRFZ7HDRWSSYTQ3P5/events.json","paper":"https://pith.science/paper/6K5VMBJY"},"agent_actions":{"view_html":"https://pith.science/pith/6K5VMBJYBWRFZ7HDRWSSYTQ3P5","download_json":"https://pith.science/pith/6K5VMBJYBWRFZ7HDRWSSYTQ3P5.json","view_paper":"https://pith.science/paper/6K5VMBJY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.12921&json=true","fetch_graph":"https://pith.science/api/pith-number/6K5VMBJYBWRFZ7HDRWSSYTQ3P5/graph.json","fetch_events":"https://pith.science/api/pith-number/6K5VMBJYBWRFZ7HDRWSSYTQ3P5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6K5VMBJYBWRFZ7HDRWSSYTQ3P5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6K5VMBJYBWRFZ7HDRWSSYTQ3P5/action/storage_attestation","attest_author":"https://pith.science/pith/6K5VMBJYBWRFZ7HDRWSSYTQ3P5/action/author_attestation","sign_citation":"https://pith.science/pith/6K5VMBJYBWRFZ7HDRWSSYTQ3P5/action/citation_signature","submit_replication":"https://pith.science/pith/6K5VMBJYBWRFZ7HDRWSSYTQ3P5/action/replication_record"}},"created_at":"2026-06-12T01:09:34.472647+00:00","updated_at":"2026-06-12T01:09:34.472647+00:00"}