{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:TQPAU7HOC7SUYB7B67RL4BT5OV","short_pith_number":"pith:TQPAU7HO","schema_version":"1.0","canonical_sha256":"9c1e0a7cee17e54c07e1f7e2be067d7550590dea5622dd297e000af8bf42dd3d","source":{"kind":"arxiv","id":"2503.22759","version":1},"attestation_state":"computed","paper":{"title":"Data Poisoning in Deep Learning: A Survey","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Di Gao, Ou Wu, Pengfei Jiao, Pinlong Zhao, Weiyao Zhu","submitted_at":"2025-03-27T15:16:57Z","abstract_excerpt":"Deep learning has become a cornerstone of modern artificial intelligence, enabling transformative applications across a wide range of domains. As the core element of deep learning, the quality and security of training data critically influence model performance and reliability. However, during the training process, deep learning models face the significant threat of data poisoning, where attackers introduce maliciously manipulated training data to degrade model accuracy or lead to anomalous behavior. While existing surveys provide valuable insights into data poisoning, they generally adopt a b"},"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":"2503.22759","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-03-27T15:16:57Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"9c752e4ae5a15b9280121c83061d2b0919a6250fe46f3da674eef14f6a6d9b54","abstract_canon_sha256":"53de7f7639ff69a2c1d6caa6e54c702997c05565e7c105b0d00b9051477fb228"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:41:25.231791Z","signature_b64":"x/0ldclAlBkZipgl8OZd10rpqE0Ea+yXOKUgwKIIfYByVFfCjIPAJdDmL8/7ADULY7WWiG/aEVlepr9I9boFAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9c1e0a7cee17e54c07e1f7e2be067d7550590dea5622dd297e000af8bf42dd3d","last_reissued_at":"2026-07-05T10:41:25.231294Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:41:25.231294Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Data Poisoning in Deep Learning: A Survey","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Di Gao, Ou Wu, Pengfei Jiao, Pinlong Zhao, Weiyao Zhu","submitted_at":"2025-03-27T15:16:57Z","abstract_excerpt":"Deep learning has become a cornerstone of modern artificial intelligence, enabling transformative applications across a wide range of domains. As the core element of deep learning, the quality and security of training data critically influence model performance and reliability. However, during the training process, deep learning models face the significant threat of data poisoning, where attackers introduce maliciously manipulated training data to degrade model accuracy or lead to anomalous behavior. While existing surveys provide valuable insights into data poisoning, they generally adopt a b"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.22759","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/2503.22759/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":"2503.22759","created_at":"2026-07-05T10:41:25.231363+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.22759v1","created_at":"2026-07-05T10:41:25.231363+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.22759","created_at":"2026-07-05T10:41:25.231363+00:00"},{"alias_kind":"pith_short_12","alias_value":"TQPAU7HOC7SU","created_at":"2026-07-05T10:41:25.231363+00:00"},{"alias_kind":"pith_short_16","alias_value":"TQPAU7HOC7SUYB7B","created_at":"2026-07-05T10:41:25.231363+00:00"},{"alias_kind":"pith_short_8","alias_value":"TQPAU7HO","created_at":"2026-07-05T10:41:25.231363+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08180","citing_title":"Out of Sight: Compression-Aware Content Protection against Agentic Crawlers","ref_index":15,"is_internal_anchor":true},{"citing_arxiv_id":"2606.12251","citing_title":"Reinforcement Learning Disrupts Gradient-Based Adversarial Optimization","ref_index":76,"is_internal_anchor":false},{"citing_arxiv_id":"2601.02947","citing_title":"Quality Degradation Attack in Synthetic Data","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2509.02372","citing_title":"Scam2Prompt: A Scalable Framework for Auditing Malicious Scam Endpoints in Production LLMs","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2604.01346","citing_title":"Safety, Security, and Cognitive Risks in World Models","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10271","citing_title":"Hijacking Text Heritage: Hiding the Human Signature through Homoglyphic Substitution","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23593","citing_title":"When AI reviews science: Can we trust the referee?","ref_index":122,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TQPAU7HOC7SUYB7B67RL4BT5OV","json":"https://pith.science/pith/TQPAU7HOC7SUYB7B67RL4BT5OV.json","graph_json":"https://pith.science/api/pith-number/TQPAU7HOC7SUYB7B67RL4BT5OV/graph.json","events_json":"https://pith.science/api/pith-number/TQPAU7HOC7SUYB7B67RL4BT5OV/events.json","paper":"https://pith.science/paper/TQPAU7HO"},"agent_actions":{"view_html":"https://pith.science/pith/TQPAU7HOC7SUYB7B67RL4BT5OV","download_json":"https://pith.science/pith/TQPAU7HOC7SUYB7B67RL4BT5OV.json","view_paper":"https://pith.science/paper/TQPAU7HO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.22759&json=true","fetch_graph":"https://pith.science/api/pith-number/TQPAU7HOC7SUYB7B67RL4BT5OV/graph.json","fetch_events":"https://pith.science/api/pith-number/TQPAU7HOC7SUYB7B67RL4BT5OV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TQPAU7HOC7SUYB7B67RL4BT5OV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TQPAU7HOC7SUYB7B67RL4BT5OV/action/storage_attestation","attest_author":"https://pith.science/pith/TQPAU7HOC7SUYB7B67RL4BT5OV/action/author_attestation","sign_citation":"https://pith.science/pith/TQPAU7HOC7SUYB7B67RL4BT5OV/action/citation_signature","submit_replication":"https://pith.science/pith/TQPAU7HOC7SUYB7B67RL4BT5OV/action/replication_record"}},"created_at":"2026-07-05T10:41:25.231363+00:00","updated_at":"2026-07-05T10:41:25.231363+00:00"}