{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:VZ6JG4S25JIX7QIOTPJBRGFV4L","short_pith_number":"pith:VZ6JG4S2","schema_version":"1.0","canonical_sha256":"ae7c93725aea517fc10e9bd21898b5e2d2c24b7a50b1e03f6a71aae16c31492a","source":{"kind":"arxiv","id":"2508.07577","version":1},"attestation_state":"computed","paper":{"title":"Exploiting Layer Normalization Fine-tuning in Visual Transformer Foundation Models for Classification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Anh Nguyen, Chen Jiang, Kai Yao, Kaizhu Huang, Qiufeng Wang, Tan Pan, Weimiao Yu, Xin Guo, Xi Yang, Yuan Cheng, Zhaorui Tan","submitted_at":"2025-08-11T03:18:47Z","abstract_excerpt":"LayerNorm is pivotal in Vision Transformers (ViTs), yet its fine-tuning dynamics under data scarcity and domain shifts remain underexplored. This paper shows that shifts in LayerNorm parameters after fine-tuning (LayerNorm shifts) are indicative of the transitions between source and target domains; its efficacy is contingent upon the degree to which the target training samples accurately represent the target domain, as quantified by our proposed Fine-tuning Shift Ratio ($FSR$). Building on this, we propose a simple yet effective rescaling mechanism using a scalar $\\lambda$ that is negatively c"},"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":"2508.07577","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-08-11T03:18:47Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"7077afa5b28c23a93153d28daa78bff7c5b715d2fa01891466158c01fd682bf8","abstract_canon_sha256":"245bb8e2c9650ad6a2f4aeeefc32a78c87d8e97eeeabfdfa0e9e4fb4dbbb3f79"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:51:48.606584Z","signature_b64":"igq1IDFazxXebNN9lluj18m64AeZFA4uw7I7ZgE4l9r1xbmkY6L/xva7zu7lI762l+xVhC75fdjyhiSn/JGsCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ae7c93725aea517fc10e9bd21898b5e2d2c24b7a50b1e03f6a71aae16c31492a","last_reissued_at":"2026-07-05T11:51:48.606101Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:51:48.606101Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Exploiting Layer Normalization Fine-tuning in Visual Transformer Foundation Models for Classification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Anh Nguyen, Chen Jiang, Kai Yao, Kaizhu Huang, Qiufeng Wang, Tan Pan, Weimiao Yu, Xin Guo, Xi Yang, Yuan Cheng, Zhaorui Tan","submitted_at":"2025-08-11T03:18:47Z","abstract_excerpt":"LayerNorm is pivotal in Vision Transformers (ViTs), yet its fine-tuning dynamics under data scarcity and domain shifts remain underexplored. This paper shows that shifts in LayerNorm parameters after fine-tuning (LayerNorm shifts) are indicative of the transitions between source and target domains; its efficacy is contingent upon the degree to which the target training samples accurately represent the target domain, as quantified by our proposed Fine-tuning Shift Ratio ($FSR$). Building on this, we propose a simple yet effective rescaling mechanism using a scalar $\\lambda$ that is negatively c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.07577","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/2508.07577/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":"2508.07577","created_at":"2026-07-05T11:51:48.606158+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.07577v1","created_at":"2026-07-05T11:51:48.606158+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.07577","created_at":"2026-07-05T11:51:48.606158+00:00"},{"alias_kind":"pith_short_12","alias_value":"VZ6JG4S25JIX","created_at":"2026-07-05T11:51:48.606158+00:00"},{"alias_kind":"pith_short_16","alias_value":"VZ6JG4S25JIX7QIO","created_at":"2026-07-05T11:51:48.606158+00:00"},{"alias_kind":"pith_short_8","alias_value":"VZ6JG4S2","created_at":"2026-07-05T11:51:48.606158+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.12766","citing_title":"PatchGen: Learning Soft Intra-Image Predictive Subsets for Visual Generalization","ref_index":231,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VZ6JG4S25JIX7QIOTPJBRGFV4L","json":"https://pith.science/pith/VZ6JG4S25JIX7QIOTPJBRGFV4L.json","graph_json":"https://pith.science/api/pith-number/VZ6JG4S25JIX7QIOTPJBRGFV4L/graph.json","events_json":"https://pith.science/api/pith-number/VZ6JG4S25JIX7QIOTPJBRGFV4L/events.json","paper":"https://pith.science/paper/VZ6JG4S2"},"agent_actions":{"view_html":"https://pith.science/pith/VZ6JG4S25JIX7QIOTPJBRGFV4L","download_json":"https://pith.science/pith/VZ6JG4S25JIX7QIOTPJBRGFV4L.json","view_paper":"https://pith.science/paper/VZ6JG4S2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.07577&json=true","fetch_graph":"https://pith.science/api/pith-number/VZ6JG4S25JIX7QIOTPJBRGFV4L/graph.json","fetch_events":"https://pith.science/api/pith-number/VZ6JG4S25JIX7QIOTPJBRGFV4L/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VZ6JG4S25JIX7QIOTPJBRGFV4L/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VZ6JG4S25JIX7QIOTPJBRGFV4L/action/storage_attestation","attest_author":"https://pith.science/pith/VZ6JG4S25JIX7QIOTPJBRGFV4L/action/author_attestation","sign_citation":"https://pith.science/pith/VZ6JG4S25JIX7QIOTPJBRGFV4L/action/citation_signature","submit_replication":"https://pith.science/pith/VZ6JG4S25JIX7QIOTPJBRGFV4L/action/replication_record"}},"created_at":"2026-07-05T11:51:48.606158+00:00","updated_at":"2026-07-05T11:51:48.606158+00:00"}