{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:Y3W2ULBWBQP5US3BVJO54KWP6O","short_pith_number":"pith:Y3W2ULBW","canonical_record":{"source":{"id":"2607.02586","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-01T05:30:07Z","cross_cats_sorted":["cs.SE"],"title_canon_sha256":"7272eee5b8a7b50fac32b884aadc35b6bff90c8cc92eed85418d92b890e1cebf","abstract_canon_sha256":"8eeef6eebdcc598930995321bc58c44b9cf41aaf0a51b69ca48d4eac5eb032b4"},"schema_version":"1.0"},"canonical_sha256":"c6edaa2c360c1fda4b61aa5dde2acff3a9681bb4ccd71b69c8eb1c15b59dc9cf","source":{"kind":"arxiv","id":"2607.02586","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.02586","created_at":"2026-07-07T00:16:01Z"},{"alias_kind":"arxiv_version","alias_value":"2607.02586v1","created_at":"2026-07-07T00:16:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.02586","created_at":"2026-07-07T00:16:01Z"},{"alias_kind":"pith_short_12","alias_value":"Y3W2ULBWBQP5","created_at":"2026-07-07T00:16:01Z"},{"alias_kind":"pith_short_16","alias_value":"Y3W2ULBWBQP5US3B","created_at":"2026-07-07T00:16:01Z"},{"alias_kind":"pith_short_8","alias_value":"Y3W2ULBW","created_at":"2026-07-07T00:16:01Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:Y3W2ULBWBQP5US3BVJO54KWP6O","target":"record","payload":{"canonical_record":{"source":{"id":"2607.02586","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-01T05:30:07Z","cross_cats_sorted":["cs.SE"],"title_canon_sha256":"7272eee5b8a7b50fac32b884aadc35b6bff90c8cc92eed85418d92b890e1cebf","abstract_canon_sha256":"8eeef6eebdcc598930995321bc58c44b9cf41aaf0a51b69ca48d4eac5eb032b4"},"schema_version":"1.0"},"canonical_sha256":"c6edaa2c360c1fda4b61aa5dde2acff3a9681bb4ccd71b69c8eb1c15b59dc9cf","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-07T00:16:01.705373Z","signature_b64":"kg+sPu24ev17prW157f3fdph1Nu/NP1LNZLkpuNLaJVTLswNNHhhahKkiFF540zkVs+R6ytryPlarMxihaOtDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c6edaa2c360c1fda4b61aa5dde2acff3a9681bb4ccd71b69c8eb1c15b59dc9cf","last_reissued_at":"2026-07-07T00:16:01.704674Z","signature_status":"signed_v1","first_computed_at":"2026-07-07T00:16:01.704674Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2607.02586","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-07T00:16:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SIoeISrGSzspXNkCWySiBwebnMboHuvDQMD7iPSKDZ1nMX35RDUbWoPruRJ5Fxl2IIw1IcS8RRFeJJznAeA0Ag==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T22:54:02.049704Z"},"content_sha256":"92b29b4f6f632525cfb551e51e56491ea4bcd21542428211fcf8b71e88a5241e","schema_version":"1.0","event_id":"sha256:92b29b4f6f632525cfb551e51e56491ea4bcd21542428211fcf8b71e88a5241e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:Y3W2ULBWBQP5US3BVJO54KWP6O","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Auditing the Audit: Five Failure Modes in Benchmark-Validity Audits","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.SE"],"primary_cat":"cs.LG","authors_text":"Yanhang Li, Zexin Zhuang, Zhichao Fan","submitted_at":"2026-07-01T05:30:07Z","abstract_excerpt":"Governance frameworks ask AI providers and auditors for documented evaluation evidence, and perturbation-based construct-validity audits are a common form of that evidence. We argue the audits are themselves fragile: their conclusions can be silently manufactured by implementation details that readers cannot see in the reported numbers. We name five classes of pipeline failure and demonstrate each in a self-audit over safety benchmarks and open-weight instruction-tuned models. Under a unified six-point due-diligence gate, every cell lands in a non-confirmatory bucket, and no cell reaches confi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.02586","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/2607.02586/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-07T00:16:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CLCAQ9lY/t7p3kAA/RU9dwdpXOix1Fam6st0MxcyGriqRH9QEIykQSU++ChsVR5euxfc0Zkjt7a9T3OcTlJ1Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T22:54:02.050198Z"},"content_sha256":"532d4026f6dcde563ea7e684c53b7a5066fc3957c355adf01c85aaf286f0c871","schema_version":"1.0","event_id":"sha256:532d4026f6dcde563ea7e684c53b7a5066fc3957c355adf01c85aaf286f0c871"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:Y3W2ULBWBQP5US3BVJO54KWP6O","target":"integrity","payload":{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.18653/v1/2025.findings-emnlp.91.URL) was visible in the surrounding text but could not be confirmed against doi.org as printed.","snippet":"URL https://aclanthology.org/2021. naacl-main.324. Lalor, J. P., Wu, H., and Yu, H. Learning latent parameters without human response patterns: Item response theory with artificial crowds. InProceedings of the 2019 Confer- ence on Empirical","arxiv_id":"2607.02586","detector":"doi_compliance","evidence":{"ref_index":2,"verdict_class":"incontrovertible","resolved_title":null,"printed_excerpt":"URL https://aclanthology.org/2021. naacl-main.324. Lalor, J. P., Wu, H., and Yu, H. Learning latent parameters without human response patterns: Item response theory with artificial crowds. InProceedings of the 2019 Confer- ence on Empirical","reconstructed_doi":"10.18653/v1/2025.findings-emnlp.91.URL"},"severity":"advisory","ref_index":2,"audited_at":"2026-07-12T11:43:17.481897Z","event_type":"pith.integrity.v1","detected_doi":"10.18653/v1/2025.findings-emnlp.91.URL","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"recoverable_identifier","evidence_hash":"9ada62cc22f217ba9d73ef34204b2b607f2b9628b2e00dff9b390418004a4a56","paper_version":1,"verdict_class":"incontrovertible","resolved_title":null,"detector_version":"1.0.0","detected_arxiv_id":null,"integrity_event_id":12601,"payload_sha256":"99d34faa3f16b22172813b1af16af99b575ee89031dc0e2b10e07edf3a0f389f","signature_b64":"dpyFBkpBmB54Pym6Ir56h8SKPZ/Wp+2dpr0RZYKHqfJXaZngaj/ge5HOKKiJFsSesPDtv44b+eyoD1LoMXQhDw==","signing_key_id":"pith-v1-2026-05"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-12T11:47:34Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"299otAVNXAZ3XincUq9TREdqKShCG1Bc6Do2nME1V+cUXlHDeFkz984oyAJSTHRZGU7TKkc/iURrH7YC5jikAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T22:54:02.054067Z"},"content_sha256":"d12ef0613e76d0b2acc6032bee7a453a4d3fd30e862890aad3d6a73acd26b157","schema_version":"1.0","event_id":"sha256:d12ef0613e76d0b2acc6032bee7a453a4d3fd30e862890aad3d6a73acd26b157"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/Y3W2ULBWBQP5US3BVJO54KWP6O/bundle.json","state_url":"https://pith.science/pith/Y3W2ULBWBQP5US3BVJO54KWP6O/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/Y3W2ULBWBQP5US3BVJO54KWP6O/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-08T22:54:02Z","links":{"resolver":"https://pith.science/pith/Y3W2ULBWBQP5US3BVJO54KWP6O","bundle":"https://pith.science/pith/Y3W2ULBWBQP5US3BVJO54KWP6O/bundle.json","state":"https://pith.science/pith/Y3W2ULBWBQP5US3BVJO54KWP6O/state.json","well_known_bundle":"https://pith.science/.well-known/pith/Y3W2ULBWBQP5US3BVJO54KWP6O/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:Y3W2ULBWBQP5US3BVJO54KWP6O","merge_version":"pith-open-graph-merge-v1","event_count":3,"valid_event_count":3,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"8eeef6eebdcc598930995321bc58c44b9cf41aaf0a51b69ca48d4eac5eb032b4","cross_cats_sorted":["cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-01T05:30:07Z","title_canon_sha256":"7272eee5b8a7b50fac32b884aadc35b6bff90c8cc92eed85418d92b890e1cebf"},"schema_version":"1.0","source":{"id":"2607.02586","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.02586","created_at":"2026-07-07T00:16:01Z"},{"alias_kind":"arxiv_version","alias_value":"2607.02586v1","created_at":"2026-07-07T00:16:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.02586","created_at":"2026-07-07T00:16:01Z"},{"alias_kind":"pith_short_12","alias_value":"Y3W2ULBWBQP5","created_at":"2026-07-07T00:16:01Z"},{"alias_kind":"pith_short_16","alias_value":"Y3W2ULBWBQP5US3B","created_at":"2026-07-07T00:16:01Z"},{"alias_kind":"pith_short_8","alias_value":"Y3W2ULBW","created_at":"2026-07-07T00:16:01Z"}],"graph_snapshots":[{"event_id":"sha256:532d4026f6dcde563ea7e684c53b7a5066fc3957c355adf01c85aaf286f0c871","target":"graph","created_at":"2026-07-07T00:16:01Z","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/2607.02586/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Governance frameworks ask AI providers and auditors for documented evaluation evidence, and perturbation-based construct-validity audits are a common form of that evidence. We argue the audits are themselves fragile: their conclusions can be silently manufactured by implementation details that readers cannot see in the reported numbers. We name five classes of pipeline failure and demonstrate each in a self-audit over safety benchmarks and open-weight instruction-tuned models. Under a unified six-point due-diligence gate, every cell lands in a non-confirmatory bucket, and no cell reaches confi","authors_text":"Yanhang Li, Zexin Zhuang, Zhichao Fan","cross_cats":["cs.SE"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-01T05:30:07Z","title":"Auditing the Audit: Five Failure Modes in Benchmark-Validity Audits"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.02586","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":"sha256:92b29b4f6f632525cfb551e51e56491ea4bcd21542428211fcf8b71e88a5241e","target":"record","created_at":"2026-07-07T00:16:01Z","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":"8eeef6eebdcc598930995321bc58c44b9cf41aaf0a51b69ca48d4eac5eb032b4","cross_cats_sorted":["cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-01T05:30:07Z","title_canon_sha256":"7272eee5b8a7b50fac32b884aadc35b6bff90c8cc92eed85418d92b890e1cebf"},"schema_version":"1.0","source":{"id":"2607.02586","kind":"arxiv","version":1}},"canonical_sha256":"c6edaa2c360c1fda4b61aa5dde2acff3a9681bb4ccd71b69c8eb1c15b59dc9cf","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c6edaa2c360c1fda4b61aa5dde2acff3a9681bb4ccd71b69c8eb1c15b59dc9cf","first_computed_at":"2026-07-07T00:16:01.704674Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-07T00:16:01.704674Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"kg+sPu24ev17prW157f3fdph1Nu/NP1LNZLkpuNLaJVTLswNNHhhahKkiFF540zkVs+R6ytryPlarMxihaOtDg==","signature_status":"signed_v1","signed_at":"2026-07-07T00:16:01.705373Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.02586","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:92b29b4f6f632525cfb551e51e56491ea4bcd21542428211fcf8b71e88a5241e","sha256:532d4026f6dcde563ea7e684c53b7a5066fc3957c355adf01c85aaf286f0c871","sha256:d12ef0613e76d0b2acc6032bee7a453a4d3fd30e862890aad3d6a73acd26b157"],"state_sha256":"01e70d4007eb3fd89e5db417d18c69ff2d5236d6fcdd533681efc7ce9789ad0d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"19yc49q82Re8KpxC4hkRWJl+ix+1Aphsja2jDhchoiRsa26OQ7mdjKttSg9T6AOjbXz/lRrIa0ITeTuVHpaKAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T22:54:02.056230Z","bundle_sha256":"842e61cc966e6568d6ebe9e912681df9e4211c86ee1808aa647972d6f587e81d"}}