{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:LEFWNJ77RDF6YOASRMCMQA5ZAJ","short_pith_number":"pith:LEFWNJ77","schema_version":"1.0","canonical_sha256":"590b66a7ff88cbec38128b04c803b902600e5df816bf4c37271b389f67903ec6","source":{"kind":"arxiv","id":"2105.07381","version":1},"attestation_state":"computed","paper":{"title":"Undistillable: Making A Nasty Teacher That CANNOT teach students","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Chenyu You, Haoyu Ma, Tianlong Chen, Ting-Kuei Hu, Xiaohui Xie, Zhangyang Wang","submitted_at":"2021-05-16T08:41:30Z","abstract_excerpt":"Knowledge Distillation (KD) is a widely used technique to transfer knowledge from pre-trained teacher models to (usually more lightweight) student models. However, in certain situations, this technique is more of a curse than a blessing. For instance, KD poses a potential risk of exposing intellectual properties (IPs): even if a trained machine learning model is released in 'black boxes' (e.g., as executable software or APIs without open-sourcing code), it can still be replicated by KD through imitating input-output behaviors. To prevent this unwanted effect of KD, this paper introduces and in"},"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":"2105.07381","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-05-16T08:41:30Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"94f27b9526cafeaa75b1ed2f5fc84ac330f06437beb2e979c5b4c81e9799c753","abstract_canon_sha256":"edc30f90c263974b1037bbc6457cc748bdf3c251ad7101b2eee2f293ced04943"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:40:40.413834Z","signature_b64":"wcy5FpYh4o31ZwrJIn+JuI9lA8H6HnrxFXspVEQqmyMj3EBHO4YAFPXJ+lDY6cbcUP1cJASCTxiSgahz+JbyBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"590b66a7ff88cbec38128b04c803b902600e5df816bf4c37271b389f67903ec6","last_reissued_at":"2026-07-05T02:40:40.413354Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:40:40.413354Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Undistillable: Making A Nasty Teacher That CANNOT teach students","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Chenyu You, Haoyu Ma, Tianlong Chen, Ting-Kuei Hu, Xiaohui Xie, Zhangyang Wang","submitted_at":"2021-05-16T08:41:30Z","abstract_excerpt":"Knowledge Distillation (KD) is a widely used technique to transfer knowledge from pre-trained teacher models to (usually more lightweight) student models. However, in certain situations, this technique is more of a curse than a blessing. For instance, KD poses a potential risk of exposing intellectual properties (IPs): even if a trained machine learning model is released in 'black boxes' (e.g., as executable software or APIs without open-sourcing code), it can still be replicated by KD through imitating input-output behaviors. To prevent this unwanted effect of KD, this paper introduces and in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2105.07381","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/2105.07381/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":"2105.07381","created_at":"2026-07-05T02:40:40.413413+00:00"},{"alias_kind":"arxiv_version","alias_value":"2105.07381v1","created_at":"2026-07-05T02:40:40.413413+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2105.07381","created_at":"2026-07-05T02:40:40.413413+00:00"},{"alias_kind":"pith_short_12","alias_value":"LEFWNJ77RDF6","created_at":"2026-07-05T02:40:40.413413+00:00"},{"alias_kind":"pith_short_16","alias_value":"LEFWNJ77RDF6YOAS","created_at":"2026-07-05T02:40:40.413413+00:00"},{"alias_kind":"pith_short_8","alias_value":"LEFWNJ77","created_at":"2026-07-05T02:40:40.413413+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.18829","citing_title":"Lossless Anti-Distillation Sampling","ref_index":100,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LEFWNJ77RDF6YOASRMCMQA5ZAJ","json":"https://pith.science/pith/LEFWNJ77RDF6YOASRMCMQA5ZAJ.json","graph_json":"https://pith.science/api/pith-number/LEFWNJ77RDF6YOASRMCMQA5ZAJ/graph.json","events_json":"https://pith.science/api/pith-number/LEFWNJ77RDF6YOASRMCMQA5ZAJ/events.json","paper":"https://pith.science/paper/LEFWNJ77"},"agent_actions":{"view_html":"https://pith.science/pith/LEFWNJ77RDF6YOASRMCMQA5ZAJ","download_json":"https://pith.science/pith/LEFWNJ77RDF6YOASRMCMQA5ZAJ.json","view_paper":"https://pith.science/paper/LEFWNJ77","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2105.07381&json=true","fetch_graph":"https://pith.science/api/pith-number/LEFWNJ77RDF6YOASRMCMQA5ZAJ/graph.json","fetch_events":"https://pith.science/api/pith-number/LEFWNJ77RDF6YOASRMCMQA5ZAJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LEFWNJ77RDF6YOASRMCMQA5ZAJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LEFWNJ77RDF6YOASRMCMQA5ZAJ/action/storage_attestation","attest_author":"https://pith.science/pith/LEFWNJ77RDF6YOASRMCMQA5ZAJ/action/author_attestation","sign_citation":"https://pith.science/pith/LEFWNJ77RDF6YOASRMCMQA5ZAJ/action/citation_signature","submit_replication":"https://pith.science/pith/LEFWNJ77RDF6YOASRMCMQA5ZAJ/action/replication_record"}},"created_at":"2026-07-05T02:40:40.413413+00:00","updated_at":"2026-07-05T02:40:40.413413+00:00"}