{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:HTU4LKIDBFPWGUOWQBUF7QN67S","short_pith_number":"pith:HTU4LKID","schema_version":"1.0","canonical_sha256":"3ce9c5a903095f6351d680685fc1befc9e5df39554e63b27b29c1bf53981a734","source":{"kind":"arxiv","id":"2403.05181","version":2},"attestation_state":"computed","paper":{"title":"Adversarial Sparse Teacher: Defense Against Distillation-Based Model Stealing Attacks Using Adversarial Examples","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CR","cs.CV"],"primary_cat":"cs.LG","authors_text":"Eda Yilmaz, Hacer Yalim Keles","submitted_at":"2024-03-08T09:43:27Z","abstract_excerpt":"We introduce Adversarial Sparse Teacher (AST), a robust defense method against distillation-based model stealing attacks. Our approach trains a teacher model using adversarial examples to produce sparse logit responses and increase the entropy of the output distribution. Typically, a model generates a peak in its output corresponding to its prediction. By leveraging adversarial examples, AST modifies the teacher model's original response, embedding a few altered logits into the output while keeping the primary response slightly higher. Concurrently, all remaining logits are elevated to further"},"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":"2403.05181","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2024-03-08T09:43:27Z","cross_cats_sorted":["cs.CR","cs.CV"],"title_canon_sha256":"46eea82a0b831070d2de105596642ba48ac9795d341c5d40b87092fc2eb140bc","abstract_canon_sha256":"0d6194a61b750cdb71e98bc664f621496a16724d2d67836248ad75fdd6cb12ac"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:46:22.557747Z","signature_b64":"1td4uXzIToPpHEH3+jD6daQbkBb1W4DMxAsmJTOey2Ffk3T7+AuUuyfjonjv8kFy12Fc11LQle6llpBoaq1KDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3ce9c5a903095f6351d680685fc1befc9e5df39554e63b27b29c1bf53981a734","last_reissued_at":"2026-07-05T08:46:22.557263Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:46:22.557263Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Adversarial Sparse Teacher: Defense Against Distillation-Based Model Stealing Attacks Using Adversarial Examples","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CR","cs.CV"],"primary_cat":"cs.LG","authors_text":"Eda Yilmaz, Hacer Yalim Keles","submitted_at":"2024-03-08T09:43:27Z","abstract_excerpt":"We introduce Adversarial Sparse Teacher (AST), a robust defense method against distillation-based model stealing attacks. Our approach trains a teacher model using adversarial examples to produce sparse logit responses and increase the entropy of the output distribution. Typically, a model generates a peak in its output corresponding to its prediction. By leveraging adversarial examples, AST modifies the teacher model's original response, embedding a few altered logits into the output while keeping the primary response slightly higher. Concurrently, all remaining logits are elevated to further"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.05181","kind":"arxiv","version":2},"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/2403.05181/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":"2403.05181","created_at":"2026-07-05T08:46:22.557323+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.05181v2","created_at":"2026-07-05T08:46:22.557323+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.05181","created_at":"2026-07-05T08:46:22.557323+00:00"},{"alias_kind":"pith_short_12","alias_value":"HTU4LKIDBFPW","created_at":"2026-07-05T08:46:22.557323+00:00"},{"alias_kind":"pith_short_16","alias_value":"HTU4LKIDBFPWGUOW","created_at":"2026-07-05T08:46:22.557323+00:00"},{"alias_kind":"pith_short_8","alias_value":"HTU4LKID","created_at":"2026-07-05T08:46:22.557323+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/HTU4LKIDBFPWGUOWQBUF7QN67S","json":"https://pith.science/pith/HTU4LKIDBFPWGUOWQBUF7QN67S.json","graph_json":"https://pith.science/api/pith-number/HTU4LKIDBFPWGUOWQBUF7QN67S/graph.json","events_json":"https://pith.science/api/pith-number/HTU4LKIDBFPWGUOWQBUF7QN67S/events.json","paper":"https://pith.science/paper/HTU4LKID"},"agent_actions":{"view_html":"https://pith.science/pith/HTU4LKIDBFPWGUOWQBUF7QN67S","download_json":"https://pith.science/pith/HTU4LKIDBFPWGUOWQBUF7QN67S.json","view_paper":"https://pith.science/paper/HTU4LKID","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.05181&json=true","fetch_graph":"https://pith.science/api/pith-number/HTU4LKIDBFPWGUOWQBUF7QN67S/graph.json","fetch_events":"https://pith.science/api/pith-number/HTU4LKIDBFPWGUOWQBUF7QN67S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HTU4LKIDBFPWGUOWQBUF7QN67S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HTU4LKIDBFPWGUOWQBUF7QN67S/action/storage_attestation","attest_author":"https://pith.science/pith/HTU4LKIDBFPWGUOWQBUF7QN67S/action/author_attestation","sign_citation":"https://pith.science/pith/HTU4LKIDBFPWGUOWQBUF7QN67S/action/citation_signature","submit_replication":"https://pith.science/pith/HTU4LKIDBFPWGUOWQBUF7QN67S/action/replication_record"}},"created_at":"2026-07-05T08:46:22.557323+00:00","updated_at":"2026-07-05T08:46:22.557323+00:00"}