{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:6WW4J4FH3BLTE25JTSPP2FCATI","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":"51a046c6a574ff8002b9d0ad1ce85ca7b2fa7d636ae3e6afee1cc5888549078b","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-26T09:43:02Z","title_canon_sha256":"196e26145a58d8b2bf92a75fd29934668b9484133efb50c67f1e89839b29191b"},"schema_version":"1.0","source":{"id":"2402.16444","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.16444","created_at":"2026-07-05T09:30:59Z"},{"alias_kind":"arxiv_version","alias_value":"2402.16444v2","created_at":"2026-07-05T09:30:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.16444","created_at":"2026-07-05T09:30:59Z"},{"alias_kind":"pith_short_12","alias_value":"6WW4J4FH3BLT","created_at":"2026-07-05T09:30:59Z"},{"alias_kind":"pith_short_16","alias_value":"6WW4J4FH3BLTE25J","created_at":"2026-07-05T09:30:59Z"},{"alias_kind":"pith_short_8","alias_value":"6WW4J4FH","created_at":"2026-07-05T09:30:59Z"}],"graph_snapshots":[{"event_id":"sha256:b8397a99253fc43fcec121bc833cbca64b346556892b503f594121995f7ba681","target":"graph","created_at":"2026-07-05T09:30:59Z","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/2402.16444/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The safety of Large Language Models (LLMs) has gained increasing attention in recent years, but there still lacks a comprehensive approach for detecting safety issues within LLMs' responses in an aligned, customizable and explainable manner. In this paper, we propose ShieldLM, an LLM-based safety detector, which aligns with common safety standards, supports customizable detection rules, and provides explanations for its decisions. To train ShieldLM, we compile a large bilingual dataset comprising 14,387 query-response pairs, annotating the safety of responses based on various safety standards.","authors_text":"Di Zhang, Hao Sun, Hongning Wang, Jingyuan Ma, Lei Sha, Minlie Huang, Pei Ke, Rui Li, Yida Lu, Zhexin Zhang, Zhifang Sui","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-26T09:43:02Z","title":"ShieldLM: Empowering LLMs as Aligned, Customizable and Explainable Safety Detectors"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.16444","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":"sha256:4c602dcf31157ef2966449f80cea2c59ad9b96ee7fd4621bc67f8a342f59da11","target":"record","created_at":"2026-07-05T09:30:59Z","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":"51a046c6a574ff8002b9d0ad1ce85ca7b2fa7d636ae3e6afee1cc5888549078b","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-26T09:43:02Z","title_canon_sha256":"196e26145a58d8b2bf92a75fd29934668b9484133efb50c67f1e89839b29191b"},"schema_version":"1.0","source":{"id":"2402.16444","kind":"arxiv","version":2}},"canonical_sha256":"f5adc4f0a7d857326ba99c9efd14409a1cb0e674cc3fa485c7ccc7579716a472","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f5adc4f0a7d857326ba99c9efd14409a1cb0e674cc3fa485c7ccc7579716a472","first_computed_at":"2026-07-05T09:30:59.744670Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:30:59.744670Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"wCylB6stesd+rnvysUCxhuFitevo9dsSY+CBjgsr0NhObH6DeQ8TtfcxqEW+hbpJaQo8vqitryFWNQYXwqNGDg==","signature_status":"signed_v1","signed_at":"2026-07-05T09:30:59.745203Z","signed_message":"canonical_sha256_bytes"},"source_id":"2402.16444","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4c602dcf31157ef2966449f80cea2c59ad9b96ee7fd4621bc67f8a342f59da11","sha256:b8397a99253fc43fcec121bc833cbca64b346556892b503f594121995f7ba681"],"state_sha256":"ef0fb47206c88ce1a0e8c1d79233cf831d96b528d4262569b5aec715997afabb"}