{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:STKZEZAMPQFOPAXVOAIPD3PW3N","short_pith_number":"pith:STKZEZAM","schema_version":"1.0","canonical_sha256":"94d592640c7c0ae782f57010f1edf6db729a327fe277f79e86522dde85c228ef","source":{"kind":"arxiv","id":"2505.22735","version":1},"attestation_state":"computed","paper":{"title":"TensorShield: Safeguarding On-Device Inference by Shielding Critical DNN Tensors with TEE","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CR","authors_text":"BoRui Li, Bowen Jiang, Hailong Lin, Tong Sun, Wei Dong, Yi Gao, Yixiao Teng","submitted_at":"2025-05-28T18:00:24Z","abstract_excerpt":"To safeguard user data privacy, on-device inference has emerged as a prominent paradigm on mobile and Internet of Things (IoT) devices. This paradigm involves deploying a model provided by a third party on local devices to perform inference tasks. However, it exposes the private model to two primary security threats: model stealing (MS) and membership inference attacks (MIA). To mitigate these risks, existing wisdom deploys models within Trusted Execution Environments (TEEs), which is a secure isolated execution space. Nonetheless, the constrained secure memory capacity in TEEs makes it challe"},"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":"2505.22735","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-05-28T18:00:24Z","cross_cats_sorted":[],"title_canon_sha256":"fb256e98fbe2f5a52dee16830309d0b23e5e5397d7216d2aad53c3dde9a58fab","abstract_canon_sha256":"9293349f39b663cc01e35f99e69c8e8a04bb54d1c0a524b50a943c7f0193d2b4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:49.324589Z","signature_b64":"QmyCJAbg2tzSGgEgBc8ZvbbtOM7GHz0DvEx3aoRyndQgVW0dOnf2ITJoBhe24/ydjp5TuOExeeB18H+354DMAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"94d592640c7c0ae782f57010f1edf6db729a327fe277f79e86522dde85c228ef","last_reissued_at":"2026-07-05T11:11:49.324072Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:49.324072Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TensorShield: Safeguarding On-Device Inference by Shielding Critical DNN Tensors with TEE","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CR","authors_text":"BoRui Li, Bowen Jiang, Hailong Lin, Tong Sun, Wei Dong, Yi Gao, Yixiao Teng","submitted_at":"2025-05-28T18:00:24Z","abstract_excerpt":"To safeguard user data privacy, on-device inference has emerged as a prominent paradigm on mobile and Internet of Things (IoT) devices. This paradigm involves deploying a model provided by a third party on local devices to perform inference tasks. However, it exposes the private model to two primary security threats: model stealing (MS) and membership inference attacks (MIA). To mitigate these risks, existing wisdom deploys models within Trusted Execution Environments (TEEs), which is a secure isolated execution space. Nonetheless, the constrained secure memory capacity in TEEs makes it challe"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.22735","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/2505.22735/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":"2505.22735","created_at":"2026-07-05T11:11:49.324136+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.22735v1","created_at":"2026-07-05T11:11:49.324136+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.22735","created_at":"2026-07-05T11:11:49.324136+00:00"},{"alias_kind":"pith_short_12","alias_value":"STKZEZAMPQFO","created_at":"2026-07-05T11:11:49.324136+00:00"},{"alias_kind":"pith_short_16","alias_value":"STKZEZAMPQFOPAXV","created_at":"2026-07-05T11:11:49.324136+00:00"},{"alias_kind":"pith_short_8","alias_value":"STKZEZAM","created_at":"2026-07-05T11:11:49.324136+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.16563","citing_title":"A Two-Stage Data Selection Framework for Data-Efficient Model Training on Edge Devices","ref_index":2025,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/STKZEZAMPQFOPAXVOAIPD3PW3N","json":"https://pith.science/pith/STKZEZAMPQFOPAXVOAIPD3PW3N.json","graph_json":"https://pith.science/api/pith-number/STKZEZAMPQFOPAXVOAIPD3PW3N/graph.json","events_json":"https://pith.science/api/pith-number/STKZEZAMPQFOPAXVOAIPD3PW3N/events.json","paper":"https://pith.science/paper/STKZEZAM"},"agent_actions":{"view_html":"https://pith.science/pith/STKZEZAMPQFOPAXVOAIPD3PW3N","download_json":"https://pith.science/pith/STKZEZAMPQFOPAXVOAIPD3PW3N.json","view_paper":"https://pith.science/paper/STKZEZAM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.22735&json=true","fetch_graph":"https://pith.science/api/pith-number/STKZEZAMPQFOPAXVOAIPD3PW3N/graph.json","fetch_events":"https://pith.science/api/pith-number/STKZEZAMPQFOPAXVOAIPD3PW3N/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/STKZEZAMPQFOPAXVOAIPD3PW3N/action/timestamp_anchor","attest_storage":"https://pith.science/pith/STKZEZAMPQFOPAXVOAIPD3PW3N/action/storage_attestation","attest_author":"https://pith.science/pith/STKZEZAMPQFOPAXVOAIPD3PW3N/action/author_attestation","sign_citation":"https://pith.science/pith/STKZEZAMPQFOPAXVOAIPD3PW3N/action/citation_signature","submit_replication":"https://pith.science/pith/STKZEZAMPQFOPAXVOAIPD3PW3N/action/replication_record"}},"created_at":"2026-07-05T11:11:49.324136+00:00","updated_at":"2026-07-05T11:11:49.324136+00:00"}