{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:L7BJTIFJU7ONC3M6IFYLRNCXOW","short_pith_number":"pith:L7BJTIFJ","schema_version":"1.0","canonical_sha256":"5fc299a0a9a7dcd16d9e4170b8b457758daa1c971dd9ba33a1f0abbd31ef2998","source":{"kind":"arxiv","id":"2501.00517","version":1},"attestation_state":"computed","paper":{"title":"A Method for Enhancing the Safety of Large Model Generation Based on Multi-dimensional Attack and Defense","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Keke Zhai","submitted_at":"2024-12-31T16:01:25Z","abstract_excerpt":"Currently, large models are prone to generating harmful content when faced with complex attack instructions, significantly reducing their defensive capabilities. To address this issue, this paper proposes a method based on constructing data aligned with multi-dimensional attack defense to enhance the generative security of large models. The core of our method lies in improving the effectiveness of safe alignment learning for large models by innova-tively increasing the diversity of attack instruction dimensions and the accuracy of generat-ing safe responses. To validate the effectiveness of ou"},"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":"2501.00517","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2024-12-31T16:01:25Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"d97f1bbc5588cadca7ead7756aa16c74fbb2a19c91e7f623de10bee42e24948c","abstract_canon_sha256":"6628ad436124df05148b90864b229ab681341f59e70c9171b8d64fcadcbce7b8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:56:07.005995Z","signature_b64":"cLl1SF6tjQfUEoi1ATknp1Q9J5aeoXIphVhSlkXQzXWX6ln5XoUFceQLD/yn9EW6yVpBEZG5QrInd27iwL3tDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5fc299a0a9a7dcd16d9e4170b8b457758daa1c971dd9ba33a1f0abbd31ef2998","last_reissued_at":"2026-07-05T09:56:07.005542Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:56:07.005542Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Method for Enhancing the Safety of Large Model Generation Based on Multi-dimensional Attack and Defense","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Keke Zhai","submitted_at":"2024-12-31T16:01:25Z","abstract_excerpt":"Currently, large models are prone to generating harmful content when faced with complex attack instructions, significantly reducing their defensive capabilities. To address this issue, this paper proposes a method based on constructing data aligned with multi-dimensional attack defense to enhance the generative security of large models. The core of our method lies in improving the effectiveness of safe alignment learning for large models by innova-tively increasing the diversity of attack instruction dimensions and the accuracy of generat-ing safe responses. To validate the effectiveness of ou"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.00517","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/2501.00517/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":"2501.00517","created_at":"2026-07-05T09:56:07.005597+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.00517v1","created_at":"2026-07-05T09:56:07.005597+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.00517","created_at":"2026-07-05T09:56:07.005597+00:00"},{"alias_kind":"pith_short_12","alias_value":"L7BJTIFJU7ON","created_at":"2026-07-05T09:56:07.005597+00:00"},{"alias_kind":"pith_short_16","alias_value":"L7BJTIFJU7ONC3M6","created_at":"2026-07-05T09:56:07.005597+00:00"},{"alias_kind":"pith_short_8","alias_value":"L7BJTIFJ","created_at":"2026-07-05T09:56:07.005597+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/L7BJTIFJU7ONC3M6IFYLRNCXOW","json":"https://pith.science/pith/L7BJTIFJU7ONC3M6IFYLRNCXOW.json","graph_json":"https://pith.science/api/pith-number/L7BJTIFJU7ONC3M6IFYLRNCXOW/graph.json","events_json":"https://pith.science/api/pith-number/L7BJTIFJU7ONC3M6IFYLRNCXOW/events.json","paper":"https://pith.science/paper/L7BJTIFJ"},"agent_actions":{"view_html":"https://pith.science/pith/L7BJTIFJU7ONC3M6IFYLRNCXOW","download_json":"https://pith.science/pith/L7BJTIFJU7ONC3M6IFYLRNCXOW.json","view_paper":"https://pith.science/paper/L7BJTIFJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.00517&json=true","fetch_graph":"https://pith.science/api/pith-number/L7BJTIFJU7ONC3M6IFYLRNCXOW/graph.json","fetch_events":"https://pith.science/api/pith-number/L7BJTIFJU7ONC3M6IFYLRNCXOW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L7BJTIFJU7ONC3M6IFYLRNCXOW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L7BJTIFJU7ONC3M6IFYLRNCXOW/action/storage_attestation","attest_author":"https://pith.science/pith/L7BJTIFJU7ONC3M6IFYLRNCXOW/action/author_attestation","sign_citation":"https://pith.science/pith/L7BJTIFJU7ONC3M6IFYLRNCXOW/action/citation_signature","submit_replication":"https://pith.science/pith/L7BJTIFJU7ONC3M6IFYLRNCXOW/action/replication_record"}},"created_at":"2026-07-05T09:56:07.005597+00:00","updated_at":"2026-07-05T09:56:07.005597+00:00"}