{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:VCTQAC3HZCA7FFRSOZFA2F46EB","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":"5dc581694927bec36b0b5796fcaeac6f1c209faadb0fd2525595422f55ee6c53","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-29T09:42:08Z","title_canon_sha256":"1ac1ab4e5752415e59c92cc880bea9c4961830302efa836ffed8d69a9c9ab080"},"schema_version":"1.0","source":{"id":"2605.31073","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.31073","created_at":"2026-06-01T01:03:34Z"},{"alias_kind":"arxiv_version","alias_value":"2605.31073v1","created_at":"2026-06-01T01:03:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.31073","created_at":"2026-06-01T01:03:34Z"},{"alias_kind":"pith_short_12","alias_value":"VCTQAC3HZCA7","created_at":"2026-06-01T01:03:34Z"},{"alias_kind":"pith_short_16","alias_value":"VCTQAC3HZCA7FFRS","created_at":"2026-06-01T01:03:34Z"},{"alias_kind":"pith_short_8","alias_value":"VCTQAC3H","created_at":"2026-06-01T01:03:34Z"}],"graph_snapshots":[{"event_id":"sha256:76922aad26669b1ca5c6e24418fab9820326d5c31fd381796a420cdc63e36fc3","target":"graph","created_at":"2026-06-01T01:03:34Z","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/2605.31073/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reasoning-based LLM guardrails improve safety moderation by generating explicit rationales before issuing final decisions. However, their rationales do not always lead to faithful enforcement: a model may recognize a harmful intent in its reasoning but still predict a safe label, or issue an unsafe decision without policy-grounded justification. We identify this safety-critical failure mode as the deliberation-to-enforcement gap. Unlike general chain-of-thought faithfulness, guardrail reliability requires policy execution consistency: the generated reasoning should be grounded in the safety po","authors_text":"Bingyu Zhu, Hui Xue, Jungang Lou, Kui Ren, Longtao Huang, Ningyu Zhang, Yan Wang, Yuefeng Chen, Zeyu Yang, Zhen Bi, Zhixuan Chu, Zihao Xue","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-29T09:42:08Z","title":"ConsisGuard: Aligning Safety Deliberation with Policy Enforcement in LLM Guardrails"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.31073","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:bffe6a5eb208fb694bebfdd5c51f6c1799a56880130133bb0efef434bbb15759","target":"record","created_at":"2026-06-01T01:03:34Z","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":"5dc581694927bec36b0b5796fcaeac6f1c209faadb0fd2525595422f55ee6c53","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-29T09:42:08Z","title_canon_sha256":"1ac1ab4e5752415e59c92cc880bea9c4961830302efa836ffed8d69a9c9ab080"},"schema_version":"1.0","source":{"id":"2605.31073","kind":"arxiv","version":1}},"canonical_sha256":"a8a7000b67c881f29632764a0d179e2055c5197902469593393e87cb58e8bdc6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a8a7000b67c881f29632764a0d179e2055c5197902469593393e87cb58e8bdc6","first_computed_at":"2026-06-01T01:03:34.639425Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-01T01:03:34.639425Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"x3NFxCJPRIHrCjry9hn1tYfrfJY7F4i+xHnIVpjJ6Saz1J0sZBqFFzynDK7xrCGEkGpt3ger3VP9ht5dSHC4Cg==","signature_status":"signed_v1","signed_at":"2026-06-01T01:03:34.640003Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.31073","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bffe6a5eb208fb694bebfdd5c51f6c1799a56880130133bb0efef434bbb15759","sha256:76922aad26669b1ca5c6e24418fab9820326d5c31fd381796a420cdc63e36fc3"],"state_sha256":"40cc2eb01597969dea30f3d2f63bcac81aaf87a9a366dd75bc3df8b4cc1e6b5b"}