{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:BF3P6B2J4EQMOXIFWVOS4OY5IA","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":"364f4575e39b3576860dcb3df39da74f23edbef7d2d1c84841df2098068b8a47","cross_cats_sorted":["cs.CR"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-26T06:54:28Z","title_canon_sha256":"49f7f9f1fd93b93b053a2f41741a3cb81f86752a57336ecb01feba8d06576ea7"},"schema_version":"1.0","source":{"id":"2607.23496","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.23496","created_at":"2026-07-28T01:22:56Z"},{"alias_kind":"arxiv_version","alias_value":"2607.23496v1","created_at":"2026-07-28T01:22:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.23496","created_at":"2026-07-28T01:22:56Z"},{"alias_kind":"pith_short_12","alias_value":"BF3P6B2J4EQM","created_at":"2026-07-28T01:22:56Z"},{"alias_kind":"pith_short_16","alias_value":"BF3P6B2J4EQMOXIF","created_at":"2026-07-28T01:22:56Z"},{"alias_kind":"pith_short_8","alias_value":"BF3P6B2J","created_at":"2026-07-28T01:22:56Z"}],"graph_snapshots":[{"event_id":"sha256:510cd2fe5ef3b22aa6853829afe8fcf6fb72e4c4f79c6fb7d77ecd9a33db1413","target":"graph","created_at":"2026-07-28T01:22:56Z","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/2607.23496/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Safety-aligned large language models are trained to refuse harmful requests, yet embedding the same requests in particular scenarios can bypass their safeguards. Existing red-teaming methods empirically identify effective scenarios through observed attack outcomes, but why particular scenarios weaken refusal remains mechanistically unclear. Meanwhile, mechanistic interpretability studies have characterized both refusal directions and jailbreak-associated features, without explaining the relationship between the two representations. In this work, we show that scenario-wrapped prompts activate i","authors_text":"Haoran Jing, Huiqi Deng, Jiahui Han, Na Zou, Xia Hu, Xiting Wang, Xuankun Rong, Ziheng Peng","cross_cats":["cs.CR"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-26T06:54:28Z","title":"Do LLMs Know Their Vulnerable Scenarios?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.23496","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:c140f0d4f9876a0a213c0ca30c097ceb559f656f8c9f39e045ddc8d3ed70e77e","target":"record","created_at":"2026-07-28T01:22:56Z","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":"364f4575e39b3576860dcb3df39da74f23edbef7d2d1c84841df2098068b8a47","cross_cats_sorted":["cs.CR"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-26T06:54:28Z","title_canon_sha256":"49f7f9f1fd93b93b053a2f41741a3cb81f86752a57336ecb01feba8d06576ea7"},"schema_version":"1.0","source":{"id":"2607.23496","kind":"arxiv","version":1}},"canonical_sha256":"0976ff0749e120c75d05b55d2e3b1d4033053fa8f39d8b38894f40d4404422ab","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0976ff0749e120c75d05b55d2e3b1d4033053fa8f39d8b38894f40d4404422ab","first_computed_at":"2026-07-28T01:22:56.047880Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-28T01:22:56.047880Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"RXy7tschq/BkTkOKBKFchY1eL21fehoR1aux7gT4hIGxUxIeDvXO/8r2ouhpuhcjjFDiy/MsJhQfqFo2bJnTCQ==","signature_status":"signed_v1","signed_at":"2026-07-28T01:22:56.048703Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.23496","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c140f0d4f9876a0a213c0ca30c097ceb559f656f8c9f39e045ddc8d3ed70e77e","sha256:510cd2fe5ef3b22aa6853829afe8fcf6fb72e4c4f79c6fb7d77ecd9a33db1413"],"state_sha256":"090bbb394e92028aff62e1c0e97f3f7615d7fbcfabea434b7b7f34230fd8fe5b"}