{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:WFFYBSCXZN5ZTLDEMVWVMXN7SZ","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":"9543f5d62d6d538520206fe3ed4bb43b6008cc73ccff29139eb55759a99d3400","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-13T17:08:42Z","title_canon_sha256":"b73851f0b8ee0e641479ab88988978766f3dfb299239a850142096b20a0cbe1e"},"schema_version":"1.0","source":{"id":"2405.07932","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.07932","created_at":"2026-07-05T08:18:53Z"},{"alias_kind":"arxiv_version","alias_value":"2405.07932v2","created_at":"2026-07-05T08:18:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.07932","created_at":"2026-07-05T08:18:53Z"},{"alias_kind":"pith_short_12","alias_value":"WFFYBSCXZN5Z","created_at":"2026-07-05T08:18:53Z"},{"alias_kind":"pith_short_16","alias_value":"WFFYBSCXZN5ZTLDE","created_at":"2026-07-05T08:18:53Z"},{"alias_kind":"pith_short_8","alias_value":"WFFYBSCX","created_at":"2026-07-05T08:18:53Z"}],"graph_snapshots":[{"event_id":"sha256:6adc84eb43cb0a5375162ad6f52f170cafdba49134841425cb4fc5727f05301a","target":"graph","created_at":"2026-07-05T08:18:53Z","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/2405.07932/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) have shown success in many natural language processing tasks. Despite rigorous safety alignment processes, supposedly safety-aligned LLMs like Llama 2 and Claude 2 are still susceptible to jailbreaks, leading to security risks and abuse of the models. One option to mitigate such risks is to augment the LLM with a dedicated \"safeguard\", which checks the LLM's inputs or outputs for undesired behaviour. A promising approach is to use the LLM itself as the safeguard. Nonetheless, baseline methods, such as prompting the LLM to self-classify toxic content, demonstrate li","authors_text":"Jakob Foerster, Qizhen Zhang, Ziyang Zhang","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-13T17:08:42Z","title":"PARDEN, Can You Repeat That? Defending against Jailbreaks via Repetition"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.07932","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":"sha256:0fc81fe002e98e0aae5454c14832320296689774a1ee63836d69af9ab5cde38a","target":"record","created_at":"2026-07-05T08:18:53Z","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":"9543f5d62d6d538520206fe3ed4bb43b6008cc73ccff29139eb55759a99d3400","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-13T17:08:42Z","title_canon_sha256":"b73851f0b8ee0e641479ab88988978766f3dfb299239a850142096b20a0cbe1e"},"schema_version":"1.0","source":{"id":"2405.07932","kind":"arxiv","version":2}},"canonical_sha256":"b14b80c857cb7b99ac64656d565dbf9649a2c6cf6a7d569466154a40bea04195","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b14b80c857cb7b99ac64656d565dbf9649a2c6cf6a7d569466154a40bea04195","first_computed_at":"2026-07-05T08:18:53.650682Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:18:53.650682Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"1p9V0iVunCXY//mVX0dDH+vt6YP4ghjC8m60JtWkiVTOJAcofst/t3Z3J7VKvr+W0K5VaYqlV/CXp5G2q3b7DA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:18:53.651197Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.07932","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0fc81fe002e98e0aae5454c14832320296689774a1ee63836d69af9ab5cde38a","sha256:6adc84eb43cb0a5375162ad6f52f170cafdba49134841425cb4fc5727f05301a"],"state_sha256":"12c88c47a8768daf70166518387b86f84fb7f16dcf71fa16d9f940b0ce8c4bf2"}