{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UWLJ7VN5JNF4AEEXDKUDFBGEYL","short_pith_number":"pith:UWLJ7VN5","schema_version":"1.0","canonical_sha256":"a5969fd5bd4b4bc010971aa83284c4c2d03f04ac696abb3b263a735ca00b6728","source":{"kind":"arxiv","id":"2607.21735","version":1},"attestation_state":"computed","paper":{"title":"What AI Red-Team Evaluations Can and Cannot Prove","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.AI","authors_text":"Bandana Kaur","submitted_at":"2026-07-23T18:34:12Z","abstract_excerpt":"Red-team evaluations of AI models support some claims and not others, and the boundary between the two is calculable rather than merely a matter of judgment. We define the evidential ceiling of an evaluation as the largest factor by which one result can move belief under a fixed testing budget, derive it in closed form for the benchmark null result, and use it to locate that boundary exactly. We find that above a calculable harm rate, a benchmark of modest size certifies a category to a stated evidentiary standard, and a clean sheet is then the stronger of the two possible observations, outwei"},"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":"2607.21735","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-23T18:34:12Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"22d0ae5ff63be9458da3be145227c92bbcc1dd9014ca3ab272f58c2ac2cfacc4","abstract_canon_sha256":"a748418608ac6900ed2829adb3b1c1c51cfe807eca6fde1d2427cb825a1ed87e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-27T00:20:18.147907Z","signature_b64":"sP023UUwWdTlyPN7ub4od3V8XLoBsA6oucvGAcAph0vVn8Kmzth6ltZjyxMWkZuDpymvVieRhFaTSKlhDBoVDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a5969fd5bd4b4bc010971aa83284c4c2d03f04ac696abb3b263a735ca00b6728","last_reissued_at":"2026-07-27T00:20:18.147083Z","signature_status":"signed_v1","first_computed_at":"2026-07-27T00:20:18.147083Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"What AI Red-Team Evaluations Can and Cannot Prove","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.AI","authors_text":"Bandana Kaur","submitted_at":"2026-07-23T18:34:12Z","abstract_excerpt":"Red-team evaluations of AI models support some claims and not others, and the boundary between the two is calculable rather than merely a matter of judgment. We define the evidential ceiling of an evaluation as the largest factor by which one result can move belief under a fixed testing budget, derive it in closed form for the benchmark null result, and use it to locate that boundary exactly. We find that above a calculable harm rate, a benchmark of modest size certifies a category to a stated evidentiary standard, and a clean sheet is then the stronger of the two possible observations, outwei"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.21735","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/2607.21735/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":"2607.21735","created_at":"2026-07-27T00:20:18.147514+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.21735v1","created_at":"2026-07-27T00:20:18.147514+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.21735","created_at":"2026-07-27T00:20:18.147514+00:00"},{"alias_kind":"pith_short_12","alias_value":"UWLJ7VN5JNF4","created_at":"2026-07-27T00:20:18.147514+00:00"},{"alias_kind":"pith_short_16","alias_value":"UWLJ7VN5JNF4AEEX","created_at":"2026-07-27T00:20:18.147514+00:00"},{"alias_kind":"pith_short_8","alias_value":"UWLJ7VN5","created_at":"2026-07-27T00:20:18.147514+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/UWLJ7VN5JNF4AEEXDKUDFBGEYL","json":"https://pith.science/pith/UWLJ7VN5JNF4AEEXDKUDFBGEYL.json","graph_json":"https://pith.science/api/pith-number/UWLJ7VN5JNF4AEEXDKUDFBGEYL/graph.json","events_json":"https://pith.science/api/pith-number/UWLJ7VN5JNF4AEEXDKUDFBGEYL/events.json","paper":"https://pith.science/paper/UWLJ7VN5"},"agent_actions":{"view_html":"https://pith.science/pith/UWLJ7VN5JNF4AEEXDKUDFBGEYL","download_json":"https://pith.science/pith/UWLJ7VN5JNF4AEEXDKUDFBGEYL.json","view_paper":"https://pith.science/paper/UWLJ7VN5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.21735&json=true","fetch_graph":"https://pith.science/api/pith-number/UWLJ7VN5JNF4AEEXDKUDFBGEYL/graph.json","fetch_events":"https://pith.science/api/pith-number/UWLJ7VN5JNF4AEEXDKUDFBGEYL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UWLJ7VN5JNF4AEEXDKUDFBGEYL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UWLJ7VN5JNF4AEEXDKUDFBGEYL/action/storage_attestation","attest_author":"https://pith.science/pith/UWLJ7VN5JNF4AEEXDKUDFBGEYL/action/author_attestation","sign_citation":"https://pith.science/pith/UWLJ7VN5JNF4AEEXDKUDFBGEYL/action/citation_signature","submit_replication":"https://pith.science/pith/UWLJ7VN5JNF4AEEXDKUDFBGEYL/action/replication_record"}},"created_at":"2026-07-27T00:20:18.147514+00:00","updated_at":"2026-07-27T00:20:18.147514+00:00"}