{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:XG5ZEQBYX3XEQKDVP66CBBLBDX","short_pith_number":"pith:XG5ZEQBY","schema_version":"1.0","canonical_sha256":"b9bb924038beee4828757fbc2085611dc405a68c0db3b36d6d5e7316aa9ba0eb","source":{"kind":"arxiv","id":"2308.04406","version":1},"attestation_state":"computed","paper":{"title":"XGBD: Explanation-Guided Graph Backdoor Detection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","cs.SI"],"primary_cat":"cs.CR","authors_text":"Mengnan Du, Ninghao Liu, Zihan Guan","submitted_at":"2023-08-08T17:10:23Z","abstract_excerpt":"Backdoor attacks pose a significant security risk to graph learning models. Backdoors can be embedded into the target model by inserting backdoor triggers into the training dataset, causing the model to make incorrect predictions when the trigger is present. To counter backdoor attacks, backdoor detection has been proposed. An emerging detection strategy in the vision and NLP domains is based on an intriguing phenomenon: when training models on a mixture of backdoor and clean samples, the loss on backdoor samples drops significantly faster than on clean samples, allowing backdoor samples to be"},"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":"2308.04406","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2023-08-08T17:10:23Z","cross_cats_sorted":["cs.LG","cs.SI"],"title_canon_sha256":"1a51b7ca8cade20341ada991a74f6420e7fed3ef4997f5c91bca5559d4452569","abstract_canon_sha256":"e8bc8db528c91fd6d68e3737931539054b54d86ab9a3f1bf25eec324c4a262bc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:39:26.968148Z","signature_b64":"9MjwZGczlRdoeSDmdw4WjMPl9A1c70mDYfO1T64AzpsmrD6N1BkJcthHiqd6qmCik95aj4paWx4shqROD6H+CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b9bb924038beee4828757fbc2085611dc405a68c0db3b36d6d5e7316aa9ba0eb","last_reissued_at":"2026-07-05T06:39:26.967741Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:39:26.967741Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"XGBD: Explanation-Guided Graph Backdoor Detection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","cs.SI"],"primary_cat":"cs.CR","authors_text":"Mengnan Du, Ninghao Liu, Zihan Guan","submitted_at":"2023-08-08T17:10:23Z","abstract_excerpt":"Backdoor attacks pose a significant security risk to graph learning models. Backdoors can be embedded into the target model by inserting backdoor triggers into the training dataset, causing the model to make incorrect predictions when the trigger is present. To counter backdoor attacks, backdoor detection has been proposed. An emerging detection strategy in the vision and NLP domains is based on an intriguing phenomenon: when training models on a mixture of backdoor and clean samples, the loss on backdoor samples drops significantly faster than on clean samples, allowing backdoor samples to be"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2308.04406","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/2308.04406/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":"2308.04406","created_at":"2026-07-05T06:39:26.967796+00:00"},{"alias_kind":"arxiv_version","alias_value":"2308.04406v1","created_at":"2026-07-05T06:39:26.967796+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2308.04406","created_at":"2026-07-05T06:39:26.967796+00:00"},{"alias_kind":"pith_short_12","alias_value":"XG5ZEQBYX3XE","created_at":"2026-07-05T06:39:26.967796+00:00"},{"alias_kind":"pith_short_16","alias_value":"XG5ZEQBYX3XEQKDV","created_at":"2026-07-05T06:39:26.967796+00:00"},{"alias_kind":"pith_short_8","alias_value":"XG5ZEQBY","created_at":"2026-07-05T06:39:26.967796+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2403.18136","citing_title":"Identifying Backdoored Graphs in Graph Neural Network Training: An Explanation-Based Approach with Novel Metrics","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XG5ZEQBYX3XEQKDVP66CBBLBDX","json":"https://pith.science/pith/XG5ZEQBYX3XEQKDVP66CBBLBDX.json","graph_json":"https://pith.science/api/pith-number/XG5ZEQBYX3XEQKDVP66CBBLBDX/graph.json","events_json":"https://pith.science/api/pith-number/XG5ZEQBYX3XEQKDVP66CBBLBDX/events.json","paper":"https://pith.science/paper/XG5ZEQBY"},"agent_actions":{"view_html":"https://pith.science/pith/XG5ZEQBYX3XEQKDVP66CBBLBDX","download_json":"https://pith.science/pith/XG5ZEQBYX3XEQKDVP66CBBLBDX.json","view_paper":"https://pith.science/paper/XG5ZEQBY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2308.04406&json=true","fetch_graph":"https://pith.science/api/pith-number/XG5ZEQBYX3XEQKDVP66CBBLBDX/graph.json","fetch_events":"https://pith.science/api/pith-number/XG5ZEQBYX3XEQKDVP66CBBLBDX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XG5ZEQBYX3XEQKDVP66CBBLBDX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XG5ZEQBYX3XEQKDVP66CBBLBDX/action/storage_attestation","attest_author":"https://pith.science/pith/XG5ZEQBYX3XEQKDVP66CBBLBDX/action/author_attestation","sign_citation":"https://pith.science/pith/XG5ZEQBYX3XEQKDVP66CBBLBDX/action/citation_signature","submit_replication":"https://pith.science/pith/XG5ZEQBYX3XEQKDVP66CBBLBDX/action/replication_record"}},"created_at":"2026-07-05T06:39:26.967796+00:00","updated_at":"2026-07-05T06:39:26.967796+00:00"}