{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:YR5PJWHLMFIOV5ETQQHOBTFCCN","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":"bd2fda9bb647d1336311159b1fd9751160fbe43c3d5bccc79b637b4541372ba7","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-17T15:54:52Z","title_canon_sha256":"3ef38260bfb8a44af3f7dd0fa6607d64b91d8238a8db40b4648a969c21b0b26e"},"schema_version":"1.0","source":{"id":"2505.12060","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.12060","created_at":"2026-07-05T11:04:57Z"},{"alias_kind":"arxiv_version","alias_value":"2505.12060v1","created_at":"2026-07-05T11:04:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12060","created_at":"2026-07-05T11:04:57Z"},{"alias_kind":"pith_short_12","alias_value":"YR5PJWHLMFIO","created_at":"2026-07-05T11:04:57Z"},{"alias_kind":"pith_short_16","alias_value":"YR5PJWHLMFIOV5ET","created_at":"2026-07-05T11:04:57Z"},{"alias_kind":"pith_short_8","alias_value":"YR5PJWHL","created_at":"2026-07-05T11:04:57Z"}],"graph_snapshots":[{"event_id":"sha256:f55dac7929d21e743259d6ccf4fb9c81c8161b08a2ddfba28458530af3b5a05d","target":"graph","created_at":"2026-07-05T11:04:57Z","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/2505.12060/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) have shown impressive capabilities across various tasks but remain vulnerable to meticulously crafted jailbreak attacks. In this paper, we identify a critical safety gap: while LLMs are adept at detecting jailbreak prompts, they often produce unsafe responses when directly processing these inputs. Inspired by this insight, we propose SAGE (Self-Aware Guard Enhancement), a training-free defense strategy designed to align LLMs' strong safety discrimination performance with their relatively weaker safety generation ability. SAGE consists of two core components: a Disc","authors_text":"Jiajun Chen, Jun Kuang, Peng Ding, Shujian Huang, Xuezhi Cao, Xunliang Cai, Zongyu Wang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-17T15:54:52Z","title":"Why Not Act on What You Know? Unleashing Safety Potential of LLMs via Self-Aware Guard Enhancement"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12060","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":"sha256:12b0f4b75fd64b139ecc58cac58fb3653a273a0d7718ce4b2a80510120df19a5","target":"record","created_at":"2026-07-05T11:04:57Z","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":"bd2fda9bb647d1336311159b1fd9751160fbe43c3d5bccc79b637b4541372ba7","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-17T15:54:52Z","title_canon_sha256":"3ef38260bfb8a44af3f7dd0fa6607d64b91d8238a8db40b4648a969c21b0b26e"},"schema_version":"1.0","source":{"id":"2505.12060","kind":"arxiv","version":1}},"canonical_sha256":"c47af4d8eb6150eaf493840ee0cca21360d6b9db6e47ce3e19082d09e7371c78","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c47af4d8eb6150eaf493840ee0cca21360d6b9db6e47ce3e19082d09e7371c78","first_computed_at":"2026-07-05T11:04:57.817147Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:04:57.817147Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hc8ScTA33J6+Lemj/K34WuA6brgd5UZhkFE0pTJAjh4ecah5eEf3xKoLekK/oAmKbeRX6yaM2QNOKCO+PWPpDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:04:57.817630Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.12060","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:12b0f4b75fd64b139ecc58cac58fb3653a273a0d7718ce4b2a80510120df19a5","sha256:f55dac7929d21e743259d6ccf4fb9c81c8161b08a2ddfba28458530af3b5a05d"],"state_sha256":"091a9c6531fcf49a24e857c74434927ed928685c74bbaacc21c1641dc528be4e"}