{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:G4VMTBZJUOSWVBEEWNF4H3PL4O","short_pith_number":"pith:G4VMTBZJ","schema_version":"1.0","canonical_sha256":"372ac98729a3a56a8484b34bc3edebe38f8428e9d4907c642c3f247f65504620","source":{"kind":"arxiv","id":"2607.16252","version":1},"attestation_state":"computed","paper":{"title":"SOS-LoRA: Static Orthogonal-Subspace Low-Rank Adaptation with Fixed Multi-Scale Scaling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Yi Chang, Yuan Wu, Yupeng Chang","submitted_at":"2026-06-27T03:25:36Z","abstract_excerpt":"Low-Rank Adaptation (LoRA) is a widely used parameter-efficient fine-tuning (PEFT) method for large language models. Under a fixed rank budget, LoRA parameterizes each adapted weight through a single low-dimensional input-side pathway, which may couple heterogeneous behaviors through shared input directions and induce interference during optimization. We propose Static Orthogonal Subspace LoRA (SOS-LoRA), a drop-in extension that reparameterizes a rank-rtot update as a sum of K static (always-on, non-routed) low-rank experts. SOS-LoRA (i) decomposes the total rank across experts, (ii) applies "},"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":"2607.16252","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-06-27T03:25:36Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"e37234ddb36229729ea138c71a5957e2af991f8c29c7f8262a1adc216886ce53","abstract_canon_sha256":"697a2e528b4998f461f74c1736a5d5aa193dac0d46560bf133bd9661cd9464fd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-21T00:20:08.368953Z","signature_b64":"L78HaCzcQybl2vwHTavgCUUNswjjzrdZeOQPhPuc2GgzUMBtlQtPxIW8PuVjZTu4lLGCYHFIy4xXZfrI9QYNDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"372ac98729a3a56a8484b34bc3edebe38f8428e9d4907c642c3f247f65504620","last_reissued_at":"2026-07-21T00:20:08.368091Z","signature_status":"signed_v1","first_computed_at":"2026-07-21T00:20:08.368091Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SOS-LoRA: Static Orthogonal-Subspace Low-Rank Adaptation with Fixed Multi-Scale Scaling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Yi Chang, Yuan Wu, Yupeng Chang","submitted_at":"2026-06-27T03:25:36Z","abstract_excerpt":"Low-Rank Adaptation (LoRA) is a widely used parameter-efficient fine-tuning (PEFT) method for large language models. Under a fixed rank budget, LoRA parameterizes each adapted weight through a single low-dimensional input-side pathway, which may couple heterogeneous behaviors through shared input directions and induce interference during optimization. We propose Static Orthogonal Subspace LoRA (SOS-LoRA), a drop-in extension that reparameterizes a rank-rtot update as a sum of K static (always-on, non-routed) low-rank experts. SOS-LoRA (i) decomposes the total rank across experts, (ii) applies "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.16252","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/2607.16252/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":"2607.16252","created_at":"2026-07-21T00:20:08.368536+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.16252v1","created_at":"2026-07-21T00:20:08.368536+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.16252","created_at":"2026-07-21T00:20:08.368536+00:00"},{"alias_kind":"pith_short_12","alias_value":"G4VMTBZJUOSW","created_at":"2026-07-21T00:20:08.368536+00:00"},{"alias_kind":"pith_short_16","alias_value":"G4VMTBZJUOSWVBEE","created_at":"2026-07-21T00:20:08.368536+00:00"},{"alias_kind":"pith_short_8","alias_value":"G4VMTBZJ","created_at":"2026-07-21T00:20:08.368536+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/G4VMTBZJUOSWVBEEWNF4H3PL4O","json":"https://pith.science/pith/G4VMTBZJUOSWVBEEWNF4H3PL4O.json","graph_json":"https://pith.science/api/pith-number/G4VMTBZJUOSWVBEEWNF4H3PL4O/graph.json","events_json":"https://pith.science/api/pith-number/G4VMTBZJUOSWVBEEWNF4H3PL4O/events.json","paper":"https://pith.science/paper/G4VMTBZJ"},"agent_actions":{"view_html":"https://pith.science/pith/G4VMTBZJUOSWVBEEWNF4H3PL4O","download_json":"https://pith.science/pith/G4VMTBZJUOSWVBEEWNF4H3PL4O.json","view_paper":"https://pith.science/paper/G4VMTBZJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.16252&json=true","fetch_graph":"https://pith.science/api/pith-number/G4VMTBZJUOSWVBEEWNF4H3PL4O/graph.json","fetch_events":"https://pith.science/api/pith-number/G4VMTBZJUOSWVBEEWNF4H3PL4O/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/G4VMTBZJUOSWVBEEWNF4H3PL4O/action/timestamp_anchor","attest_storage":"https://pith.science/pith/G4VMTBZJUOSWVBEEWNF4H3PL4O/action/storage_attestation","attest_author":"https://pith.science/pith/G4VMTBZJUOSWVBEEWNF4H3PL4O/action/author_attestation","sign_citation":"https://pith.science/pith/G4VMTBZJUOSWVBEEWNF4H3PL4O/action/citation_signature","submit_replication":"https://pith.science/pith/G4VMTBZJUOSWVBEEWNF4H3PL4O/action/replication_record"}},"created_at":"2026-07-21T00:20:08.368536+00:00","updated_at":"2026-07-21T00:20:08.368536+00:00"}