{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:3J2YGJQ3T43FIV7MJQKH4LYLMP","short_pith_number":"pith:3J2YGJQ3","schema_version":"1.0","canonical_sha256":"da7583261b9f365457ec4c147e2f0b63f8288085d742acc6821d36e8c4366ceb","source":{"kind":"arxiv","id":"2503.17198","version":1},"attestation_state":"computed","paper":{"title":"Jailbreaking the Non-Transferable Barrier via Test-Time Data Disguising","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"cs.CR","authors_text":"Dadong Wang, Lina Yao, Tongliang Liu, Yongli Xiang, Ziming Hong","submitted_at":"2025-03-21T14:47:33Z","abstract_excerpt":"Non-transferable learning (NTL) has been proposed to protect model intellectual property (IP) by creating a \"non-transferable barrier\" to restrict generalization from authorized to unauthorized domains. Recently, well-designed attack, which restores the unauthorized-domain performance by fine-tuning NTL models on few authorized samples, highlights the security risks of NTL-based applications. However, such attack requires modifying model weights, thus being invalid in the black-box scenario. This raises a critical question: can we trust the security of NTL models deployed as black-box systems?"},"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":"2503.17198","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-03-21T14:47:33Z","cross_cats_sorted":["cs.CV","cs.LG"],"title_canon_sha256":"1abadc211b0e3fc683d71f8f2446546a641f7dcf02864bc5203b7f8a42284c80","abstract_canon_sha256":"086660237d124d5c507cefe717c77563074e13990c085d3bf288db3317f908cd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:36:59.880920Z","signature_b64":"zeTXGkErwJEV4VGzpruta9qXGBs8ZiNZOLWrUC6QbFjR2fx3FEOwd2XJIFqg3iH4l9EIRFRResgKY2WN3XViAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"da7583261b9f365457ec4c147e2f0b63f8288085d742acc6821d36e8c4366ceb","last_reissued_at":"2026-07-05T10:36:59.880442Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:36:59.880442Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Jailbreaking the Non-Transferable Barrier via Test-Time Data Disguising","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"cs.CR","authors_text":"Dadong Wang, Lina Yao, Tongliang Liu, Yongli Xiang, Ziming Hong","submitted_at":"2025-03-21T14:47:33Z","abstract_excerpt":"Non-transferable learning (NTL) has been proposed to protect model intellectual property (IP) by creating a \"non-transferable barrier\" to restrict generalization from authorized to unauthorized domains. Recently, well-designed attack, which restores the unauthorized-domain performance by fine-tuning NTL models on few authorized samples, highlights the security risks of NTL-based applications. However, such attack requires modifying model weights, thus being invalid in the black-box scenario. This raises a critical question: can we trust the security of NTL models deployed as black-box systems?"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.17198","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/2503.17198/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":"2503.17198","created_at":"2026-07-05T10:36:59.880495+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.17198v1","created_at":"2026-07-05T10:36:59.880495+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.17198","created_at":"2026-07-05T10:36:59.880495+00:00"},{"alias_kind":"pith_short_12","alias_value":"3J2YGJQ3T43F","created_at":"2026-07-05T10:36:59.880495+00:00"},{"alias_kind":"pith_short_16","alias_value":"3J2YGJQ3T43FIV7M","created_at":"2026-07-05T10:36:59.880495+00:00"},{"alias_kind":"pith_short_8","alias_value":"3J2YGJQ3","created_at":"2026-07-05T10:36:59.880495+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.04119","citing_title":"When Data-Free Knowledge Distillation Meets Non-Transferable Teacher: Escaping Out-of-Distribution Trap is All You Need","ref_index":18,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3J2YGJQ3T43FIV7MJQKH4LYLMP","json":"https://pith.science/pith/3J2YGJQ3T43FIV7MJQKH4LYLMP.json","graph_json":"https://pith.science/api/pith-number/3J2YGJQ3T43FIV7MJQKH4LYLMP/graph.json","events_json":"https://pith.science/api/pith-number/3J2YGJQ3T43FIV7MJQKH4LYLMP/events.json","paper":"https://pith.science/paper/3J2YGJQ3"},"agent_actions":{"view_html":"https://pith.science/pith/3J2YGJQ3T43FIV7MJQKH4LYLMP","download_json":"https://pith.science/pith/3J2YGJQ3T43FIV7MJQKH4LYLMP.json","view_paper":"https://pith.science/paper/3J2YGJQ3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.17198&json=true","fetch_graph":"https://pith.science/api/pith-number/3J2YGJQ3T43FIV7MJQKH4LYLMP/graph.json","fetch_events":"https://pith.science/api/pith-number/3J2YGJQ3T43FIV7MJQKH4LYLMP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3J2YGJQ3T43FIV7MJQKH4LYLMP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3J2YGJQ3T43FIV7MJQKH4LYLMP/action/storage_attestation","attest_author":"https://pith.science/pith/3J2YGJQ3T43FIV7MJQKH4LYLMP/action/author_attestation","sign_citation":"https://pith.science/pith/3J2YGJQ3T43FIV7MJQKH4LYLMP/action/citation_signature","submit_replication":"https://pith.science/pith/3J2YGJQ3T43FIV7MJQKH4LYLMP/action/replication_record"}},"created_at":"2026-07-05T10:36:59.880495+00:00","updated_at":"2026-07-05T10:36:59.880495+00:00"}