{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:JOMRU4EDVT5KNMZVROKAXVSN7M","short_pith_number":"pith:JOMRU4ED","canonical_record":{"source":{"id":"2406.01288","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T12:59:17Z","cross_cats_sorted":["cs.AI","cs.CR","cs.LG"],"title_canon_sha256":"f5cb188eb0d926d1f89a581abcc7cb0ddc7f1a97649cf8a85bbfd2c5ecd507a8","abstract_canon_sha256":"9e88f7daa79d0c81bd9c087887a4dc12de43b7af4fe9b13ae5802c71190663e9"},"schema_version":"1.0"},"canonical_sha256":"4b991a7083acfaa6b3358b940bd64dfb3fd2ba581617135d80c67c204ba6243f","source":{"kind":"arxiv","id":"2406.01288","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.01288","created_at":"2026-07-05T09:28:31Z"},{"alias_kind":"arxiv_version","alias_value":"2406.01288v2","created_at":"2026-07-05T09:28:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.01288","created_at":"2026-07-05T09:28:31Z"},{"alias_kind":"pith_short_12","alias_value":"JOMRU4EDVT5K","created_at":"2026-07-05T09:28:31Z"},{"alias_kind":"pith_short_16","alias_value":"JOMRU4EDVT5KNMZV","created_at":"2026-07-05T09:28:31Z"},{"alias_kind":"pith_short_8","alias_value":"JOMRU4ED","created_at":"2026-07-05T09:28:31Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:JOMRU4EDVT5KNMZVROKAXVSN7M","target":"record","payload":{"canonical_record":{"source":{"id":"2406.01288","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T12:59:17Z","cross_cats_sorted":["cs.AI","cs.CR","cs.LG"],"title_canon_sha256":"f5cb188eb0d926d1f89a581abcc7cb0ddc7f1a97649cf8a85bbfd2c5ecd507a8","abstract_canon_sha256":"9e88f7daa79d0c81bd9c087887a4dc12de43b7af4fe9b13ae5802c71190663e9"},"schema_version":"1.0"},"canonical_sha256":"4b991a7083acfaa6b3358b940bd64dfb3fd2ba581617135d80c67c204ba6243f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:28:31.744904Z","signature_b64":"ZT257j28Jy5Tabt401aLVY7c4hcMMNVC04jD2cqzA92VFoK/CmlBZ6O4Gjhfha1+LjmD8ii0IEMTfKJGREo/Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4b991a7083acfaa6b3358b940bd64dfb3fd2ba581617135d80c67c204ba6243f","last_reissued_at":"2026-07-05T09:28:31.744342Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:28:31.744342Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.01288","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":"2026-07-05T09:28:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ArOsmXOZSs41D0KtKQmOvqUqyue84l4mNfpK1ylmFnHtzqUM12YE5L0497X8sZaGebqAl6o0oO48mrLTaco4DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-10T09:47:30.846636Z"},"content_sha256":"44339fce38c49b7c51ec22fe42b4179ca72c7c415e9be67f4e8adb6ef39ffa71","schema_version":"1.0","event_id":"sha256:44339fce38c49b7c51ec22fe42b4179ca72c7c415e9be67f4e8adb6ef39ffa71"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:JOMRU4EDVT5KNMZVROKAXVSN7M","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Improved Few-Shot Jailbreaking Can Circumvent Aligned Language Models and Their Defenses","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CR","cs.LG"],"primary_cat":"cs.CL","authors_text":"Chao Du, Jing Jiang, Min Lin, Qian Liu, Tianyu Pang, Xiaosen Zheng","submitted_at":"2024-06-03T12:59:17Z","abstract_excerpt":"Recently, Anil et al. (2024) show that many-shot (up to hundreds of) demonstrations can jailbreak state-of-the-art LLMs by exploiting their long-context capability. Nevertheless, is it possible to use few-shot demonstrations to efficiently jailbreak LLMs within limited context sizes? While the vanilla few-shot jailbreaking may be inefficient, we propose improved techniques such as injecting special system tokens like [/INST] and employing demo-level random search from a collected demo pool. These simple techniques result in surprisingly effective jailbreaking against aligned LLMs (even with ad"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.01288","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/2406.01288/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":"2026-07-05T09:28:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PWSnCejmF52daQqanZKYQkZV+vzyo2X1n8usBRpuZZ4cManpwm1XBzjtzQr6bKxx6XE6buaqlRU5Ef7D31CXAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-10T09:47:30.847031Z"},"content_sha256":"bbc962cace8241be3be3fe9ffb40ccd27eb27c60fde65f20e5e1aa3434696a2f","schema_version":"1.0","event_id":"sha256:bbc962cace8241be3be3fe9ffb40ccd27eb27c60fde65f20e5e1aa3434696a2f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JOMRU4EDVT5KNMZVROKAXVSN7M/bundle.json","state_url":"https://pith.science/pith/JOMRU4EDVT5KNMZVROKAXVSN7M/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JOMRU4EDVT5KNMZVROKAXVSN7M/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":"2026-07-10T09:47:30Z","links":{"resolver":"https://pith.science/pith/JOMRU4EDVT5KNMZVROKAXVSN7M","bundle":"https://pith.science/pith/JOMRU4EDVT5KNMZVROKAXVSN7M/bundle.json","state":"https://pith.science/pith/JOMRU4EDVT5KNMZVROKAXVSN7M/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JOMRU4EDVT5KNMZVROKAXVSN7M/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:JOMRU4EDVT5KNMZVROKAXVSN7M","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":"9e88f7daa79d0c81bd9c087887a4dc12de43b7af4fe9b13ae5802c71190663e9","cross_cats_sorted":["cs.AI","cs.CR","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T12:59:17Z","title_canon_sha256":"f5cb188eb0d926d1f89a581abcc7cb0ddc7f1a97649cf8a85bbfd2c5ecd507a8"},"schema_version":"1.0","source":{"id":"2406.01288","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.01288","created_at":"2026-07-05T09:28:31Z"},{"alias_kind":"arxiv_version","alias_value":"2406.01288v2","created_at":"2026-07-05T09:28:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.01288","created_at":"2026-07-05T09:28:31Z"},{"alias_kind":"pith_short_12","alias_value":"JOMRU4EDVT5K","created_at":"2026-07-05T09:28:31Z"},{"alias_kind":"pith_short_16","alias_value":"JOMRU4EDVT5KNMZV","created_at":"2026-07-05T09:28:31Z"},{"alias_kind":"pith_short_8","alias_value":"JOMRU4ED","created_at":"2026-07-05T09:28:31Z"}],"graph_snapshots":[{"event_id":"sha256:bbc962cace8241be3be3fe9ffb40ccd27eb27c60fde65f20e5e1aa3434696a2f","target":"graph","created_at":"2026-07-05T09:28:31Z","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/2406.01288/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recently, Anil et al. (2024) show that many-shot (up to hundreds of) demonstrations can jailbreak state-of-the-art LLMs by exploiting their long-context capability. Nevertheless, is it possible to use few-shot demonstrations to efficiently jailbreak LLMs within limited context sizes? While the vanilla few-shot jailbreaking may be inefficient, we propose improved techniques such as injecting special system tokens like [/INST] and employing demo-level random search from a collected demo pool. These simple techniques result in surprisingly effective jailbreaking against aligned LLMs (even with ad","authors_text":"Chao Du, Jing Jiang, Min Lin, Qian Liu, Tianyu Pang, Xiaosen Zheng","cross_cats":["cs.AI","cs.CR","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T12:59:17Z","title":"Improved Few-Shot Jailbreaking Can Circumvent Aligned Language Models and Their Defenses"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.01288","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":"sha256:44339fce38c49b7c51ec22fe42b4179ca72c7c415e9be67f4e8adb6ef39ffa71","target":"record","created_at":"2026-07-05T09:28:31Z","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":"9e88f7daa79d0c81bd9c087887a4dc12de43b7af4fe9b13ae5802c71190663e9","cross_cats_sorted":["cs.AI","cs.CR","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T12:59:17Z","title_canon_sha256":"f5cb188eb0d926d1f89a581abcc7cb0ddc7f1a97649cf8a85bbfd2c5ecd507a8"},"schema_version":"1.0","source":{"id":"2406.01288","kind":"arxiv","version":2}},"canonical_sha256":"4b991a7083acfaa6b3358b940bd64dfb3fd2ba581617135d80c67c204ba6243f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4b991a7083acfaa6b3358b940bd64dfb3fd2ba581617135d80c67c204ba6243f","first_computed_at":"2026-07-05T09:28:31.744342Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:28:31.744342Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ZT257j28Jy5Tabt401aLVY7c4hcMMNVC04jD2cqzA92VFoK/CmlBZ6O4Gjhfha1+LjmD8ii0IEMTfKJGREo/Cg==","signature_status":"signed_v1","signed_at":"2026-07-05T09:28:31.744904Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.01288","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:44339fce38c49b7c51ec22fe42b4179ca72c7c415e9be67f4e8adb6ef39ffa71","sha256:bbc962cace8241be3be3fe9ffb40ccd27eb27c60fde65f20e5e1aa3434696a2f"],"state_sha256":"3de1ed78a74b135432916c18792bc31030cfc0e3c5eaf32f4fefd6fc6b0395d4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9yBJJ/UJLs4BMnpeaGXLwc9ORotvf8zE/uXkO2ytm/lLTZED4NwtmE9HMiY3RWpBps16JUcIsD2kDBwFjMlCAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-10T09:47:30.848997Z","bundle_sha256":"8f88e5e9c3c0b44f41282e603786910bc9fa558ae7db9ee7be5710c5aaf8be65"}}