{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:QWB2XN47WU63ZS4RVJXYFVPJX6","short_pith_number":"pith:QWB2XN47","canonical_record":{"source":{"id":"2407.15399","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-22T06:04:29Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"06e44d49e32027516d518a169d7a492db561885e6a6e01481ef5c91750a4f1f4","abstract_canon_sha256":"1e389af73ee2ead53fb35c448497bb52b8c2a2aa3b06300c5b76cc57488a2488"},"schema_version":"1.0"},"canonical_sha256":"8583abb79fb53dbccb91aa6f82d5e9bfb3588491743ce71c24fb9db14bb5021f","source":{"kind":"arxiv","id":"2407.15399","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.15399","created_at":"2026-07-05T08:46:43Z"},{"alias_kind":"arxiv_version","alias_value":"2407.15399v1","created_at":"2026-07-05T08:46:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.15399","created_at":"2026-07-05T08:46:43Z"},{"alias_kind":"pith_short_12","alias_value":"QWB2XN47WU63","created_at":"2026-07-05T08:46:43Z"},{"alias_kind":"pith_short_16","alias_value":"QWB2XN47WU63ZS4R","created_at":"2026-07-05T08:46:43Z"},{"alias_kind":"pith_short_8","alias_value":"QWB2XN47","created_at":"2026-07-05T08:46:43Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:QWB2XN47WU63ZS4RVJXYFVPJX6","target":"record","payload":{"canonical_record":{"source":{"id":"2407.15399","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-22T06:04:29Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"06e44d49e32027516d518a169d7a492db561885e6a6e01481ef5c91750a4f1f4","abstract_canon_sha256":"1e389af73ee2ead53fb35c448497bb52b8c2a2aa3b06300c5b76cc57488a2488"},"schema_version":"1.0"},"canonical_sha256":"8583abb79fb53dbccb91aa6f82d5e9bfb3588491743ce71c24fb9db14bb5021f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:46:43.232709Z","signature_b64":"+BSrrcbFRlP4O/OchIVB26563aTclFhaz6zL9VICCdIkzCkdsWc9VkduigU8jEHRijMveYg2BxzM87kwLRt6Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8583abb79fb53dbccb91aa6f82d5e9bfb3588491743ce71c24fb9db14bb5021f","last_reissued_at":"2026-07-05T08:46:43.232239Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:46:43.232239Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2407.15399","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:46:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nSnlJHKLWcRhJJCSs21RkIlUK8UOmSLHx0UtKOz3Lmrb/0oY5MZQPsDqb9hzAgYHArKslGXdMdqwdRpKX5qoDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T05:17:08.059675Z"},"content_sha256":"672d54392ee74c40b69422bbb5e81590ea3495d30150e1cf20c34ea7c982e36c","schema_version":"1.0","event_id":"sha256:672d54392ee74c40b69422bbb5e81590ea3495d30150e1cf20c34ea7c982e36c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:QWB2XN47WU63ZS4RVJXYFVPJX6","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Imposter.AI: Adversarial Attacks with Hidden Intentions towards Aligned Large Language Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.CL","authors_text":"Fuying Ye, Liangzhi Li, Lu wei, Noa Garcia, Tong Xiang, Wangyue Li, Xiao Liu","submitted_at":"2024-07-22T06:04:29Z","abstract_excerpt":"With the development of large language models (LLMs) like ChatGPT, both their vast applications and potential vulnerabilities have come to the forefront. While developers have integrated multiple safety mechanisms to mitigate their misuse, a risk remains, particularly when models encounter adversarial inputs. This study unveils an attack mechanism that capitalizes on human conversation strategies to extract harmful information from LLMs. We delineate three pivotal strategies: (i) decomposing malicious questions into seemingly innocent sub-questions; (ii) rewriting overtly malicious questions i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.15399","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/2407.15399/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":"2026-07-05T08:46:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gKQomNpx5fDWIQzoV73V9oJ2G/zvTUnWNOL7linFzKAKsW2doX9LzUm+n6FbAevOig9/1cYhSBWaFIuBYUz/Cg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T05:17:08.060579Z"},"content_sha256":"35430f9e20fe0e1b131dce4291840d93505b8f0f2d621a56dc3257227e4d4673","schema_version":"1.0","event_id":"sha256:35430f9e20fe0e1b131dce4291840d93505b8f0f2d621a56dc3257227e4d4673"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/QWB2XN47WU63ZS4RVJXYFVPJX6/bundle.json","state_url":"https://pith.science/pith/QWB2XN47WU63ZS4RVJXYFVPJX6/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/QWB2XN47WU63ZS4RVJXYFVPJX6/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":"2026-08-10T05:17:08Z","links":{"resolver":"https://pith.science/pith/QWB2XN47WU63ZS4RVJXYFVPJX6","bundle":"https://pith.science/pith/QWB2XN47WU63ZS4RVJXYFVPJX6/bundle.json","state":"https://pith.science/pith/QWB2XN47WU63ZS4RVJXYFVPJX6/state.json","well_known_bundle":"https://pith.science/.well-known/pith/QWB2XN47WU63ZS4RVJXYFVPJX6/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:QWB2XN47WU63ZS4RVJXYFVPJX6","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":"1e389af73ee2ead53fb35c448497bb52b8c2a2aa3b06300c5b76cc57488a2488","cross_cats_sorted":["cs.AI","cs.CR"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-22T06:04:29Z","title_canon_sha256":"06e44d49e32027516d518a169d7a492db561885e6a6e01481ef5c91750a4f1f4"},"schema_version":"1.0","source":{"id":"2407.15399","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.15399","created_at":"2026-07-05T08:46:43Z"},{"alias_kind":"arxiv_version","alias_value":"2407.15399v1","created_at":"2026-07-05T08:46:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.15399","created_at":"2026-07-05T08:46:43Z"},{"alias_kind":"pith_short_12","alias_value":"QWB2XN47WU63","created_at":"2026-07-05T08:46:43Z"},{"alias_kind":"pith_short_16","alias_value":"QWB2XN47WU63ZS4R","created_at":"2026-07-05T08:46:43Z"},{"alias_kind":"pith_short_8","alias_value":"QWB2XN47","created_at":"2026-07-05T08:46:43Z"}],"graph_snapshots":[{"event_id":"sha256:35430f9e20fe0e1b131dce4291840d93505b8f0f2d621a56dc3257227e4d4673","target":"graph","created_at":"2026-07-05T08:46:43Z","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/2407.15399/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"With the development of large language models (LLMs) like ChatGPT, both their vast applications and potential vulnerabilities have come to the forefront. While developers have integrated multiple safety mechanisms to mitigate their misuse, a risk remains, particularly when models encounter adversarial inputs. This study unveils an attack mechanism that capitalizes on human conversation strategies to extract harmful information from LLMs. We delineate three pivotal strategies: (i) decomposing malicious questions into seemingly innocent sub-questions; (ii) rewriting overtly malicious questions i","authors_text":"Fuying Ye, Liangzhi Li, Lu wei, Noa Garcia, Tong Xiang, Wangyue Li, Xiao Liu","cross_cats":["cs.AI","cs.CR"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-22T06:04:29Z","title":"Imposter.AI: Adversarial Attacks with Hidden Intentions towards Aligned Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.15399","kind":"arxiv","version":1},"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":"sha256:672d54392ee74c40b69422bbb5e81590ea3495d30150e1cf20c34ea7c982e36c","target":"record","created_at":"2026-07-05T08:46:43Z","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":"1e389af73ee2ead53fb35c448497bb52b8c2a2aa3b06300c5b76cc57488a2488","cross_cats_sorted":["cs.AI","cs.CR"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-22T06:04:29Z","title_canon_sha256":"06e44d49e32027516d518a169d7a492db561885e6a6e01481ef5c91750a4f1f4"},"schema_version":"1.0","source":{"id":"2407.15399","kind":"arxiv","version":1}},"canonical_sha256":"8583abb79fb53dbccb91aa6f82d5e9bfb3588491743ce71c24fb9db14bb5021f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8583abb79fb53dbccb91aa6f82d5e9bfb3588491743ce71c24fb9db14bb5021f","first_computed_at":"2026-07-05T08:46:43.232239Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:46:43.232239Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"+BSrrcbFRlP4O/OchIVB26563aTclFhaz6zL9VICCdIkzCkdsWc9VkduigU8jEHRijMveYg2BxzM87kwLRt6Ag==","signature_status":"signed_v1","signed_at":"2026-07-05T08:46:43.232709Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.15399","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:672d54392ee74c40b69422bbb5e81590ea3495d30150e1cf20c34ea7c982e36c","sha256:35430f9e20fe0e1b131dce4291840d93505b8f0f2d621a56dc3257227e4d4673"],"state_sha256":"3fbae0f685141ba049b1bad0826db0e80542fde91341f903b3bd680e14f25da3"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IbVwS0SeF5LrkIlPoGzlg4pf0oQ2oi7/Af4r7w3HiNlaJwFcdNsmxvA/56hwRV1RXkmsxbwfoXv74VtLiW2eAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T05:17:08.065662Z","bundle_sha256":"62571dc9d6421427e8680ef14be46476c644b6523d4c544ab063828b9234a933"}}