{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:OQKJOKDQWB53XUNPZ5EYWUVUZ7","short_pith_number":"pith:OQKJOKDQ","schema_version":"1.0","canonical_sha256":"7414972870b07bbbd1afcf498b52b4cfdc6501a2018b167233470a65a21441e7","source":{"kind":"arxiv","id":"2606.10456","version":1},"attestation_state":"computed","paper":{"title":"The Distributed Detectability Band Against Marginal-Preserving Attacks","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Gao Yuze, Zhang Qinqin","submitted_at":"2026-06-09T06:04:30Z","abstract_excerpt":"AI-control monitors score individual agent actions to detect misbehavior, but real harm can be distributed across many benign-looking steps, each individually below any per-step alarm. We construct a marginal-preserving, correlation-encoded distributed-sabotage attack using a Gaussian-copula AR(1) construction: the per-step monitor-score marginal is held exactly equal to benign, so mean, max, top-k tail, and threshold monitors (Monitor A) are defeated by construction, while harm is encoded in the temporal correlation structure. We sequence the paper around three reviewer-mandated gates. (1) Re"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2606.10456","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CR","submitted_at":"2026-06-09T06:04:30Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"502fb9e7584fb33c1be304817a25b58d51ca71f1363cd553d9b14a77374d69f7","abstract_canon_sha256":"5a1025cbba1538639e8044f58952ceb94f740cd10389b73fe934cc31c776eda0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-10T01:10:20.147830Z","signature_b64":"/h4TsWMkIUwUYSaZpahtYHuLpHN5T2a6Jd9IIOJOThdpNqgzQaWFCtJlIW6C5JS7AF7WlW5Uhe8FGJYD0AvnBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7414972870b07bbbd1afcf498b52b4cfdc6501a2018b167233470a65a21441e7","last_reissued_at":"2026-06-10T01:10:20.146917Z","signature_status":"signed_v1","first_computed_at":"2026-06-10T01:10:20.146917Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Distributed Detectability Band Against Marginal-Preserving Attacks","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Gao Yuze, Zhang Qinqin","submitted_at":"2026-06-09T06:04:30Z","abstract_excerpt":"AI-control monitors score individual agent actions to detect misbehavior, but real harm can be distributed across many benign-looking steps, each individually below any per-step alarm. We construct a marginal-preserving, correlation-encoded distributed-sabotage attack using a Gaussian-copula AR(1) construction: the per-step monitor-score marginal is held exactly equal to benign, so mean, max, top-k tail, and threshold monitors (Monitor A) are defeated by construction, while harm is encoded in the temporal correlation structure. We sequence the paper around three reviewer-mandated gates. (1) Re"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.10456","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2606.10456/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2606.10456","created_at":"2026-06-10T01:10:20.147084+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.10456v1","created_at":"2026-06-10T01:10:20.147084+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.10456","created_at":"2026-06-10T01:10:20.147084+00:00"},{"alias_kind":"pith_short_12","alias_value":"OQKJOKDQWB53","created_at":"2026-06-10T01:10:20.147084+00:00"},{"alias_kind":"pith_short_16","alias_value":"OQKJOKDQWB53XUNP","created_at":"2026-06-10T01:10:20.147084+00:00"},{"alias_kind":"pith_short_8","alias_value":"OQKJOKDQ","created_at":"2026-06-10T01:10:20.147084+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OQKJOKDQWB53XUNPZ5EYWUVUZ7","json":"https://pith.science/pith/OQKJOKDQWB53XUNPZ5EYWUVUZ7.json","graph_json":"https://pith.science/api/pith-number/OQKJOKDQWB53XUNPZ5EYWUVUZ7/graph.json","events_json":"https://pith.science/api/pith-number/OQKJOKDQWB53XUNPZ5EYWUVUZ7/events.json","paper":"https://pith.science/paper/OQKJOKDQ"},"agent_actions":{"view_html":"https://pith.science/pith/OQKJOKDQWB53XUNPZ5EYWUVUZ7","download_json":"https://pith.science/pith/OQKJOKDQWB53XUNPZ5EYWUVUZ7.json","view_paper":"https://pith.science/paper/OQKJOKDQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.10456&json=true","fetch_graph":"https://pith.science/api/pith-number/OQKJOKDQWB53XUNPZ5EYWUVUZ7/graph.json","fetch_events":"https://pith.science/api/pith-number/OQKJOKDQWB53XUNPZ5EYWUVUZ7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OQKJOKDQWB53XUNPZ5EYWUVUZ7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OQKJOKDQWB53XUNPZ5EYWUVUZ7/action/storage_attestation","attest_author":"https://pith.science/pith/OQKJOKDQWB53XUNPZ5EYWUVUZ7/action/author_attestation","sign_citation":"https://pith.science/pith/OQKJOKDQWB53XUNPZ5EYWUVUZ7/action/citation_signature","submit_replication":"https://pith.science/pith/OQKJOKDQWB53XUNPZ5EYWUVUZ7/action/replication_record"}},"created_at":"2026-06-10T01:10:20.147084+00:00","updated_at":"2026-06-10T01:10:20.147084+00:00"}