{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:QPCPLFWW6ZZWSBGJJ5UA7SNCE3","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":"ad7de2a6a426d93bbc32c830e6e17e7574b995113123ed0b2d5408e2de46bde6","cross_cats_sorted":["cs.SE"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-20T18:08:21Z","title_canon_sha256":"cc97ddd8d2f188556db2e6e5098adf58220df7cc5228dbbb3929d920c5019aa0"},"schema_version":"1.0","source":{"id":"2605.21602","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.21602","created_at":"2026-05-22T01:03:24Z"},{"alias_kind":"arxiv_version","alias_value":"2605.21602v1","created_at":"2026-05-22T01:03:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.21602","created_at":"2026-05-22T01:03:24Z"},{"alias_kind":"pith_short_12","alias_value":"QPCPLFWW6ZZW","created_at":"2026-05-22T01:03:24Z"},{"alias_kind":"pith_short_16","alias_value":"QPCPLFWW6ZZWSBGJ","created_at":"2026-05-22T01:03:24Z"},{"alias_kind":"pith_short_8","alias_value":"QPCPLFWW","created_at":"2026-05-22T01:03:24Z"}],"graph_snapshots":[{"event_id":"sha256:30e046ca7379e25bd34bd53141b6a24edafd9429341e436b677094fa49b5bee2","target":"graph","created_at":"2026-05-22T01:03:24Z","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/2605.21602/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Many safety and alignment failures of large language models (LLMs) occur due to out-of-distribution (OOD) situations: unusual prompt or response patterns that are unforeseen by model developers. We systematically study whether LLM monitoring pipelines can detect these OOD alignment failures by introducing a benchmark called Misalignment Out Of Distribution (MOOD). It is difficult to find failures that are truly OOD for off-the-shelf models trained on vast safety datasets. We sidestep this by including a restricted training set in MOOD that we use to train our own monitors, as well as seven tes","authors_text":"Cassidy Laidlaw, Dylan Feng, Pragya Srivastava","cross_cats":["cs.SE"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-20T18:08:21Z","title":"Benchmarking and Improving Monitors for Out-Of-Distribution Alignment Failure in LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.21602","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:d36f75553085313da0be18f20a594c2360a046a07f7d98dfb63e8cd1cc4f376f","target":"record","created_at":"2026-05-22T01:03:24Z","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":"ad7de2a6a426d93bbc32c830e6e17e7574b995113123ed0b2d5408e2de46bde6","cross_cats_sorted":["cs.SE"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-20T18:08:21Z","title_canon_sha256":"cc97ddd8d2f188556db2e6e5098adf58220df7cc5228dbbb3929d920c5019aa0"},"schema_version":"1.0","source":{"id":"2605.21602","kind":"arxiv","version":1}},"canonical_sha256":"83c4f596d6f6736904c94f680fc9a226f1431fd883133d3118f2010bb2bcfc2d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"83c4f596d6f6736904c94f680fc9a226f1431fd883133d3118f2010bb2bcfc2d","first_computed_at":"2026-05-22T01:03:24.865833Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-22T01:03:24.865833Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"uQN0dWEt2dmznhF05vjzHKDFfcV1CpkOuKAFBTN8e1e0RkiaLiY3OuBvILIG+DwM3WrZaFHE6LLAbDPJ9nRyAg==","signature_status":"signed_v1","signed_at":"2026-05-22T01:03:24.866311Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.21602","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d36f75553085313da0be18f20a594c2360a046a07f7d98dfb63e8cd1cc4f376f","sha256:30e046ca7379e25bd34bd53141b6a24edafd9429341e436b677094fa49b5bee2"],"state_sha256":"5e9e3b835479be61dc711f81e4e0c75a17e210ef0ee984b77d43db1b05061d48"}