{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:SIDUSORETCLFGXXTOAKTLDII66","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":"92c856c173e8db7a34331bd57e12fca46be3d6bb6866d7fbb8081796a322fd63","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.HC","submitted_at":"2025-05-21T10:56:16Z","title_canon_sha256":"b729190ca1718248ef46b92b37c167bd6f550d2e3fc7909074082ee84ba8bf4d"},"schema_version":"1.0","source":{"id":"2505.15365","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.15365","created_at":"2026-07-05T11:06:44Z"},{"alias_kind":"arxiv_version","alias_value":"2505.15365v1","created_at":"2026-07-05T11:06:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.15365","created_at":"2026-07-05T11:06:44Z"},{"alias_kind":"pith_short_12","alias_value":"SIDUSORETCLF","created_at":"2026-07-05T11:06:44Z"},{"alias_kind":"pith_short_16","alias_value":"SIDUSORETCLFGXXT","created_at":"2026-07-05T11:06:44Z"},{"alias_kind":"pith_short_8","alias_value":"SIDUSORE","created_at":"2026-07-05T11:06:44Z"}],"graph_snapshots":[{"event_id":"sha256:81283cc2523468dd17c81a4bb47bb71ffe31d6d7b2fb65e4786b1786172a8322","target":"graph","created_at":"2026-07-05T11:06:44Z","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/2505.15365/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As large language models (LLMs) are increasingly deployed in high-stakes settings, their ability to refuse ethically sensitive prompts-such as those involving hate speech or illegal activities-has become central to content moderation and responsible AI practices. While refusal responses can be viewed as evidence of ethical alignment and safety-conscious behavior, recent research suggests that users may perceive them negatively. At the same time, automated assessments of model outputs are playing a growing role in both evaluation and training. In particular, LLM-as-a-Judge frameworks-in which o","authors_text":"Stefan Pasch","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.HC","submitted_at":"2025-05-21T10:56:16Z","title":"AI vs. Human Judgment of Content Moderation: LLM-as-a-Judge and Ethics-Based Response Refusals"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.15365","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:e188481e29cbdc61452d9f689a0063b22b56b03d17239973fd845cc332ddff9f","target":"record","created_at":"2026-07-05T11:06:44Z","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":"92c856c173e8db7a34331bd57e12fca46be3d6bb6866d7fbb8081796a322fd63","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.HC","submitted_at":"2025-05-21T10:56:16Z","title_canon_sha256":"b729190ca1718248ef46b92b37c167bd6f550d2e3fc7909074082ee84ba8bf4d"},"schema_version":"1.0","source":{"id":"2505.15365","kind":"arxiv","version":1}},"canonical_sha256":"9207493a249896535ef37015358d08f7be3e262ab31524891413b405b048c5e6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9207493a249896535ef37015358d08f7be3e262ab31524891413b405b048c5e6","first_computed_at":"2026-07-05T11:06:44.305559Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:06:44.305559Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"b1wZ9HhADz/Tk0DcuuMM/t4k3T07ITVxB3G4RBi169gbgZ7+tbQIM7q5y1BlXzbzJjHWxyQvDjpr/7zhYS/QCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:06:44.306060Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.15365","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e188481e29cbdc61452d9f689a0063b22b56b03d17239973fd845cc332ddff9f","sha256:81283cc2523468dd17c81a4bb47bb71ffe31d6d7b2fb65e4786b1786172a8322"],"state_sha256":"d52a38d629720c92ff585f0ec20e380a5b30928754249b508a278302cec92205"}