{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:4CRWGGVF476MI52JK6MUBSABHF","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":"b49e792501d1e8a2c9dab8842c915f9d87a16640a250aa46af45fd39ba7702f3","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-05T22:11:26Z","title_canon_sha256":"0d14ae625adf38eb42fb57dc56f88cc89ca99e0ba726625b246a5ce592a740c5"},"schema_version":"1.0","source":{"id":"2606.07874","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.07874","created_at":"2026-06-09T01:04:54Z"},{"alias_kind":"arxiv_version","alias_value":"2606.07874v1","created_at":"2026-06-09T01:04:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.07874","created_at":"2026-06-09T01:04:54Z"},{"alias_kind":"pith_short_12","alias_value":"4CRWGGVF476M","created_at":"2026-06-09T01:04:54Z"},{"alias_kind":"pith_short_16","alias_value":"4CRWGGVF476MI52J","created_at":"2026-06-09T01:04:54Z"},{"alias_kind":"pith_short_8","alias_value":"4CRWGGVF","created_at":"2026-06-09T01:04:54Z"}],"graph_snapshots":[{"event_id":"sha256:73e1949db69b535b7dbd5b67f93f021a7101d57612fb42d91e3a08979c4185d8","target":"graph","created_at":"2026-06-09T01:04:54Z","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/2606.07874/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"LLMs-as-judges are the only way to evaluate safety at scale. Despite their importance, LLM-judges themselves are rarely evaluated beyond human agreement in simple, static benchmarks. We therefore investigate two under-explored but crucial properties of LLMs-as-judges: their susceptibility to relying on in context-information, and their steerability to differing safety definitions, which may not align with their internal safety priors. We evaluate the safety judging abilities of many generalist LLMs and safety-specific judges, and investigate the impact of task demonstrations, novel in-context ","authors_text":"Anissa Alloula, Ava Batchkala, Federico Licini, Seraphina Goldfarb-Tarrant","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-05T22:11:26Z","title":"Safety is Contextual, LLM-Judges Are Not: Navigating the Rigid Priors of Evaluators"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.07874","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":"sha256:722cd55614e72674ad1dbe8da25902fc548ab7b9fa116258ef5320a40662f772","target":"record","created_at":"2026-06-09T01:04:54Z","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":"b49e792501d1e8a2c9dab8842c915f9d87a16640a250aa46af45fd39ba7702f3","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-05T22:11:26Z","title_canon_sha256":"0d14ae625adf38eb42fb57dc56f88cc89ca99e0ba726625b246a5ce592a740c5"},"schema_version":"1.0","source":{"id":"2606.07874","kind":"arxiv","version":1}},"canonical_sha256":"e0a3631aa5e7fcc47749579940c80139629c57d09f2a62828c83bdd507663b79","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e0a3631aa5e7fcc47749579940c80139629c57d09f2a62828c83bdd507663b79","first_computed_at":"2026-06-09T01:04:54.287492Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-09T01:04:54.287492Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"X+jRXDS6takH0vj34l67d8sxES9LdtQjrqGN65kPYz5uKJ5W3gpUatdZ+AYus6VzA86yaafaknPFoqa5Y0HAAQ==","signature_status":"signed_v1","signed_at":"2026-06-09T01:04:54.287859Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.07874","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:722cd55614e72674ad1dbe8da25902fc548ab7b9fa116258ef5320a40662f772","sha256:73e1949db69b535b7dbd5b67f93f021a7101d57612fb42d91e3a08979c4185d8"],"state_sha256":"ff70034a535261b79777e6245ef842d9a840960776bf4e0b00d8b006e9a903c2"}