{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:7K5SDFW4TEWNMR3RESA3GNLNUZ","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":"78b8dd90a764453c74b03f260af7f204dc7fd4047473671f3d7a999436a1219c","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-28T08:40:32Z","title_canon_sha256":"fcd2bf5845c63e6a81744dc67fc39897865ad456e6ae634e985104821242a6d2"},"schema_version":"1.0","source":{"id":"2605.29601","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.29601","created_at":"2026-05-29T01:05:50Z"},{"alias_kind":"arxiv_version","alias_value":"2605.29601v1","created_at":"2026-05-29T01:05:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.29601","created_at":"2026-05-29T01:05:50Z"},{"alias_kind":"pith_short_12","alias_value":"7K5SDFW4TEWN","created_at":"2026-05-29T01:05:50Z"},{"alias_kind":"pith_short_16","alias_value":"7K5SDFW4TEWNMR3R","created_at":"2026-05-29T01:05:50Z"},{"alias_kind":"pith_short_8","alias_value":"7K5SDFW4","created_at":"2026-05-29T01:05:50Z"}],"graph_snapshots":[{"event_id":"sha256:910f6e51271ee3c052423b9f1cee70480410024889f28896a73c20a39fdba4f4","target":"graph","created_at":"2026-05-29T01:05:50Z","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/2605.29601/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As autonomous agents become more capable of performing real-world tasks, distinguishing scheming behavior from benign task pursuit may become a central AI control problem. Existing monitors often rely on chain-of-thought access or internal activations, or use prompted frontier models, all of which can be unavailable, unreliable or expensive in deployment. In this work, we study action-only deliberative monitors: smaller open-weight models trained to detect scheming and sabotage from agentic trajectories without accessing the monitored agent's reasoning or model internals. Our method, inspired ","authors_text":"Aditya Sinha, Akshat Naik, Axel H{\\o}jmark, Kilian Merkelbach, Marius Hobbhahn, Rich Barton-Cooper, Simon Storf, Victor Gillioz","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-28T08:40:32Z","title":"Training Deliberative Monitors for Black-Box Scheming Detection"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.29601","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":"sha256:3df4ccb8501735c80e7ea26f2b053ce8be564e78685572ea8bff0cece45c5837","target":"record","created_at":"2026-05-29T01:05:50Z","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":"78b8dd90a764453c74b03f260af7f204dc7fd4047473671f3d7a999436a1219c","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-28T08:40:32Z","title_canon_sha256":"fcd2bf5845c63e6a81744dc67fc39897865ad456e6ae634e985104821242a6d2"},"schema_version":"1.0","source":{"id":"2605.29601","kind":"arxiv","version":1}},"canonical_sha256":"fabb2196dc992cd647712481b3356da66a5c148cc26902dd100ebb20c07b45f3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"fabb2196dc992cd647712481b3356da66a5c148cc26902dd100ebb20c07b45f3","first_computed_at":"2026-05-29T01:05:50.342923Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-29T01:05:50.342923Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"fS+9dT6Fxf8RjxU2R21KoplSIeqP4ZwxBTkUVpYO5JT1ktEZZItr3eqo6SM/gS37LVcdjw2PgEjUgX1Tb+w7Ag==","signature_status":"signed_v1","signed_at":"2026-05-29T01:05:50.343526Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.29601","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3df4ccb8501735c80e7ea26f2b053ce8be564e78685572ea8bff0cece45c5837","sha256:910f6e51271ee3c052423b9f1cee70480410024889f28896a73c20a39fdba4f4"],"state_sha256":"b1b3dca6ff34154e35e42a50d70ac70d59c9dbd456965632a959fec7b6305255"}