{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:O7SQFOW7XOKUYFBVPYWHGAQTLJ","short_pith_number":"pith:O7SQFOW7","canonical_record":{"source":{"id":"2507.06043","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-07-08T14:45:21Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"595d92510be167272fc32045bb7230612dd4e94bdaaf1e427e4f6f94d68f1b7a","abstract_canon_sha256":"6530dae7980803ba77eee873b1f07d898cd22e5119206905043e3d39c6aee725"},"schema_version":"1.0"},"canonical_sha256":"77e502badfbb954c14357e2c7302135a4bdb5e854272acf21ac08d7dad9b1f5d","source":{"kind":"arxiv","id":"2507.06043","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.06043","created_at":"2026-07-05T11:49:11Z"},{"alias_kind":"arxiv_version","alias_value":"2507.06043v2","created_at":"2026-07-05T11:49:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.06043","created_at":"2026-07-05T11:49:11Z"},{"alias_kind":"pith_short_12","alias_value":"O7SQFOW7XOKU","created_at":"2026-07-05T11:49:11Z"},{"alias_kind":"pith_short_16","alias_value":"O7SQFOW7XOKUYFBV","created_at":"2026-07-05T11:49:11Z"},{"alias_kind":"pith_short_8","alias_value":"O7SQFOW7","created_at":"2026-07-05T11:49:11Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:O7SQFOW7XOKUYFBVPYWHGAQTLJ","target":"record","payload":{"canonical_record":{"source":{"id":"2507.06043","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-07-08T14:45:21Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"595d92510be167272fc32045bb7230612dd4e94bdaaf1e427e4f6f94d68f1b7a","abstract_canon_sha256":"6530dae7980803ba77eee873b1f07d898cd22e5119206905043e3d39c6aee725"},"schema_version":"1.0"},"canonical_sha256":"77e502badfbb954c14357e2c7302135a4bdb5e854272acf21ac08d7dad9b1f5d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:49:11.284604Z","signature_b64":"PXWIVVNCVmi4ap4lg5Fe1ccVgVcjl4eHfGL4wV4zw6rVxjv17yySjZ4TCOs+u1RZ12vSZH1oXuFPRkDHKPykBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"77e502badfbb954c14357e2c7302135a4bdb5e854272acf21ac08d7dad9b1f5d","last_reissued_at":"2026-07-05T11:49:11.284057Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:49:11.284057Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2507.06043","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:49:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5T5/v/69iaya2cKcHXlCj+slJUvNqjqfBRBrQP0VSwTBfl3uCZqaBeCr1Whc6NA5YqgDrUUXVBE1ImO/NzN4Cg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T04:49:03.404641Z"},"content_sha256":"eba7ae25be344cf40f69f2f40b66273f27017520a48bc8ccd3722869d209f2ca","schema_version":"1.0","event_id":"sha256:eba7ae25be344cf40f69f2f40b66273f27017520a48bc8ccd3722869d209f2ca"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:O7SQFOW7XOKUYFBVPYWHGAQTLJ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"CAVGAN: Unifying Jailbreak and Defense of LLMs via Generative Adversarial Attacks on their Internal Representations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Jianhao Chen, Mayi Xu, Tieyun Qian, Xiaohu Li, Yunfeng Ning, Zepeng Bao","submitted_at":"2025-07-08T14:45:21Z","abstract_excerpt":"Security alignment enables the Large Language Model (LLM) to gain the protection against malicious queries, but various jailbreak attack methods reveal the vulnerability of this security mechanism. Previous studies have isolated LLM jailbreak attacks and defenses. We analyze the security protection mechanism of the LLM, and propose a framework that combines attack and defense. Our method is based on the linearly separable property of LLM intermediate layer embedding, as well as the essence of jailbreak attack, which aims to embed harmful problems and transfer them to the safe area. We utilize "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.06043","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/2507.06043/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:49:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lLLx9hW8Exdr1+UKbeSAgcSgR1OWRlrpLEiyMcJ099oUSSwD98u/Pq1XDLZ/f+put+mxFTOcEH5A3x1cBeXjDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T04:49:03.405202Z"},"content_sha256":"bc1ef2fd299726a4d65501fdf927b8201871fad8c9bbc39c8142310ce71eb11c","schema_version":"1.0","event_id":"sha256:bc1ef2fd299726a4d65501fdf927b8201871fad8c9bbc39c8142310ce71eb11c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/O7SQFOW7XOKUYFBVPYWHGAQTLJ/bundle.json","state_url":"https://pith.science/pith/O7SQFOW7XOKUYFBVPYWHGAQTLJ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/O7SQFOW7XOKUYFBVPYWHGAQTLJ/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-07T04:49:03Z","links":{"resolver":"https://pith.science/pith/O7SQFOW7XOKUYFBVPYWHGAQTLJ","bundle":"https://pith.science/pith/O7SQFOW7XOKUYFBVPYWHGAQTLJ/bundle.json","state":"https://pith.science/pith/O7SQFOW7XOKUYFBVPYWHGAQTLJ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/O7SQFOW7XOKUYFBVPYWHGAQTLJ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:O7SQFOW7XOKUYFBVPYWHGAQTLJ","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"6530dae7980803ba77eee873b1f07d898cd22e5119206905043e3d39c6aee725","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-07-08T14:45:21Z","title_canon_sha256":"595d92510be167272fc32045bb7230612dd4e94bdaaf1e427e4f6f94d68f1b7a"},"schema_version":"1.0","source":{"id":"2507.06043","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.06043","created_at":"2026-07-05T11:49:11Z"},{"alias_kind":"arxiv_version","alias_value":"2507.06043v2","created_at":"2026-07-05T11:49:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.06043","created_at":"2026-07-05T11:49:11Z"},{"alias_kind":"pith_short_12","alias_value":"O7SQFOW7XOKU","created_at":"2026-07-05T11:49:11Z"},{"alias_kind":"pith_short_16","alias_value":"O7SQFOW7XOKUYFBV","created_at":"2026-07-05T11:49:11Z"},{"alias_kind":"pith_short_8","alias_value":"O7SQFOW7","created_at":"2026-07-05T11:49:11Z"}],"graph_snapshots":[{"event_id":"sha256:bc1ef2fd299726a4d65501fdf927b8201871fad8c9bbc39c8142310ce71eb11c","target":"graph","created_at":"2026-07-05T11:49:11Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2507.06043/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Security alignment enables the Large Language Model (LLM) to gain the protection against malicious queries, but various jailbreak attack methods reveal the vulnerability of this security mechanism. Previous studies have isolated LLM jailbreak attacks and defenses. We analyze the security protection mechanism of the LLM, and propose a framework that combines attack and defense. Our method is based on the linearly separable property of LLM intermediate layer embedding, as well as the essence of jailbreak attack, which aims to embed harmful problems and transfer them to the safe area. We utilize ","authors_text":"Jianhao Chen, Mayi Xu, Tieyun Qian, Xiaohu Li, Yunfeng Ning, Zepeng Bao","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-07-08T14:45:21Z","title":"CAVGAN: Unifying Jailbreak and Defense of LLMs via Generative Adversarial Attacks on their Internal Representations"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.06043","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:eba7ae25be344cf40f69f2f40b66273f27017520a48bc8ccd3722869d209f2ca","target":"record","created_at":"2026-07-05T11:49:11Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"6530dae7980803ba77eee873b1f07d898cd22e5119206905043e3d39c6aee725","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-07-08T14:45:21Z","title_canon_sha256":"595d92510be167272fc32045bb7230612dd4e94bdaaf1e427e4f6f94d68f1b7a"},"schema_version":"1.0","source":{"id":"2507.06043","kind":"arxiv","version":2}},"canonical_sha256":"77e502badfbb954c14357e2c7302135a4bdb5e854272acf21ac08d7dad9b1f5d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"77e502badfbb954c14357e2c7302135a4bdb5e854272acf21ac08d7dad9b1f5d","first_computed_at":"2026-07-05T11:49:11.284057Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:49:11.284057Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"PXWIVVNCVmi4ap4lg5Fe1ccVgVcjl4eHfGL4wV4zw6rVxjv17yySjZ4TCOs+u1RZ12vSZH1oXuFPRkDHKPykBw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:49:11.284604Z","signed_message":"canonical_sha256_bytes"},"source_id":"2507.06043","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:eba7ae25be344cf40f69f2f40b66273f27017520a48bc8ccd3722869d209f2ca","sha256:bc1ef2fd299726a4d65501fdf927b8201871fad8c9bbc39c8142310ce71eb11c"],"state_sha256":"806ce3aa73fad82da700e320f00a171bd40207e6e09940dcc5bcb81e9863c719"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NiawB6/xdnIbmVeuoxPZBfEw0SEqlq1JGh+YhtXdVwEhGP4EIRU6bSg+SzylaytBTkaTT+QZAYNrmuJtQuLLCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T04:49:03.409478Z","bundle_sha256":"41401049b4924fb722c7ec458aa392cee15e8d38d490e1d5f45054ca1f45ea5f"}}