{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:NG5QIF47MSEUVG4FYB34NER6HH","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":"4b404f5320196e6302122d0dfa40aeee1ba87c11a573e2cce0f87f9e8c4e252e","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-07-03T16:03:10Z","title_canon_sha256":"62435e10220147c0ed9baa0fa052c2194b927d226a02f640a549bbf6a33ce457"},"schema_version":"1.0","source":{"id":"2407.03232","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.03232","created_at":"2026-07-05T08:39:48Z"},{"alias_kind":"arxiv_version","alias_value":"2407.03232v1","created_at":"2026-07-05T08:39:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.03232","created_at":"2026-07-05T08:39:48Z"},{"alias_kind":"pith_short_12","alias_value":"NG5QIF47MSEU","created_at":"2026-07-05T08:39:48Z"},{"alias_kind":"pith_short_16","alias_value":"NG5QIF47MSEUVG4F","created_at":"2026-07-05T08:39:48Z"},{"alias_kind":"pith_short_8","alias_value":"NG5QIF47","created_at":"2026-07-05T08:39:48Z"}],"graph_snapshots":[{"event_id":"sha256:0d0d31f81fce6e4b588d82ee65cb01d9a226aa43510c4b8b524dd4dc7be447ed","target":"graph","created_at":"2026-07-05T08:39:48Z","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/2407.03232/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"When LLMs are deployed in sensitive, human-facing settings, it is crucial that they do not output unsafe, biased, or privacy-violating outputs. For this reason, models are both trained and instructed to refuse to answer unsafe prompts such as \"Tell me how to build a bomb.\" We find that, despite these safeguards, it is possible to break model defenses simply by appending a space to the end of a model's input. In a study of eight open-source models, we demonstrate that this acts as a strong enough attack to cause the majority of models to generate harmful outputs with very high success rates. We","authors_text":"Hannah Brown, Kenji Kawaguchi, Leon Lin, Michael Shieh","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-07-03T16:03:10Z","title":"Single Character Perturbations Break LLM Alignment"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.03232","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":"sha256:4ee199e3b345aad529975cfb3aaffd0c59c736ae195e49177b3be26becc71f7b","target":"record","created_at":"2026-07-05T08:39:48Z","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":"4b404f5320196e6302122d0dfa40aeee1ba87c11a573e2cce0f87f9e8c4e252e","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-07-03T16:03:10Z","title_canon_sha256":"62435e10220147c0ed9baa0fa052c2194b927d226a02f640a549bbf6a33ce457"},"schema_version":"1.0","source":{"id":"2407.03232","kind":"arxiv","version":1}},"canonical_sha256":"69bb04179f64894a9b85c077c6923e39e762c60311d5959e8c52ada9ceb41723","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"69bb04179f64894a9b85c077c6923e39e762c60311d5959e8c52ada9ceb41723","first_computed_at":"2026-07-05T08:39:48.054077Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:39:48.054077Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"55BgDEwI71s1h9qX38o7vKwevqNvXf0HljeX4verz5lho/p4vDwcfsT5/taG7w77lZ6/XKxcwUBMtaC+btpGAw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:39:48.054512Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.03232","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4ee199e3b345aad529975cfb3aaffd0c59c736ae195e49177b3be26becc71f7b","sha256:0d0d31f81fce6e4b588d82ee65cb01d9a226aa43510c4b8b524dd4dc7be447ed"],"state_sha256":"c876340241c5bd3dcaa747566199768056ca4398812f0e487f30fd9e884d74e7"}