{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:BN56XD2IMCJTN4CCJG25ZNR2PQ","short_pith_number":"pith:BN56XD2I","canonical_record":{"source":{"id":"2608.02786","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-03T18:33:37Z","cross_cats_sorted":[],"title_canon_sha256":"802206483ed5c71bfa269c24acc1fa179f352a73220af924998525e3f89f0f76","abstract_canon_sha256":"58840026ef55daca0497db17725ef912b8631c87dc2cbd3cf3e8a531d690fe70"},"schema_version":"1.0"},"canonical_sha256":"0b7beb8f48609336f04249b5dcb63a7c353c98b22b11873392d84e4e53ff44d0","source":{"kind":"arxiv","id":"2608.02786","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.02786","created_at":"2026-08-05T00:40:02Z"},{"alias_kind":"arxiv_version","alias_value":"2608.02786v1","created_at":"2026-08-05T00:40:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.02786","created_at":"2026-08-05T00:40:02Z"},{"alias_kind":"pith_short_12","alias_value":"BN56XD2IMCJT","created_at":"2026-08-05T00:40:02Z"},{"alias_kind":"pith_short_16","alias_value":"BN56XD2IMCJTN4CC","created_at":"2026-08-05T00:40:02Z"},{"alias_kind":"pith_short_8","alias_value":"BN56XD2I","created_at":"2026-08-05T00:40:02Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:BN56XD2IMCJTN4CCJG25ZNR2PQ","target":"record","payload":{"canonical_record":{"source":{"id":"2608.02786","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-03T18:33:37Z","cross_cats_sorted":[],"title_canon_sha256":"802206483ed5c71bfa269c24acc1fa179f352a73220af924998525e3f89f0f76","abstract_canon_sha256":"58840026ef55daca0497db17725ef912b8631c87dc2cbd3cf3e8a531d690fe70"},"schema_version":"1.0"},"canonical_sha256":"0b7beb8f48609336f04249b5dcb63a7c353c98b22b11873392d84e4e53ff44d0","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-05T00:40:02.224004Z","signature_b64":"Mlo/dxXcMF+bYO0LoeBSL3/k1fSprZxldDjibuDLrppziZ1JqksRBmCLyu3pmU8kW3tpkQawaqDyKC2ok1QDBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0b7beb8f48609336f04249b5dcb63a7c353c98b22b11873392d84e4e53ff44d0","last_reissued_at":"2026-08-05T00:40:02.220996Z","signature_status":"signed_v1","first_computed_at":"2026-08-05T00:40:02.220996Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2608.02786","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":"2026-08-05T00:40:02Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Dyfo//zfpY8oX/bWCHmKrAIJHvzwCVpwFpyoCjtYnkJsTKgYTNgBgiWWpJwzj6y9KW8MYj13B+JBKg076cl5BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-16T05:23:52.678749Z"},"content_sha256":"94b9056de1b3c936a6578adee4cac71dbc0faaaf02dfb18c517616029f076f9e","schema_version":"1.0","event_id":"sha256:94b9056de1b3c936a6578adee4cac71dbc0faaaf02dfb18c517616029f076f9e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:BN56XD2IMCJTN4CCJG25ZNR2PQ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Evaluation Blindness: How Silent Measurement Failures Corrupt AI Systems from Training to Deployment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Priyanka Bajaj (Independent Researcher)","submitted_at":"2026-08-03T18:33:37Z","abstract_excerpt":"AI systems can fail silently. The failure propagates through training loops, evaluation pipelines, and production monitoring stacks until downstream harm makes it visible. This paper introduces evaluation blindness: a measurement function M exhibits evaluation blindness with respect to failure class F when it produces readings indistinguishable from a healthy state while the system is actually failing, with no auxiliary signal flagging the gap.\n  The problem surfaces at two lifecycle stages the literature has treated separately. At training time, reward models are gamed, importance-sampling co"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.02786","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/2608.02786/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":"2026-08-05T00:40:02Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"JiVqNvs8BaE6+I+bKuuKXBBABpYOJS/zeNSXjTTuAB5P1iQv0A5wBdgowLt1/foob73cChzNMcPLk0HAgj96Ag==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-16T05:23:52.679261Z"},"content_sha256":"b6589088bf925b9ca5f6de7ba90c9453f9dcc347e93a344a5969ff5548807245","schema_version":"1.0","event_id":"sha256:b6589088bf925b9ca5f6de7ba90c9453f9dcc347e93a344a5969ff5548807245"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/BN56XD2IMCJTN4CCJG25ZNR2PQ/bundle.json","state_url":"https://pith.science/pith/BN56XD2IMCJTN4CCJG25ZNR2PQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/BN56XD2IMCJTN4CCJG25ZNR2PQ/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":"2026-08-16T05:23:52Z","links":{"resolver":"https://pith.science/pith/BN56XD2IMCJTN4CCJG25ZNR2PQ","bundle":"https://pith.science/pith/BN56XD2IMCJTN4CCJG25ZNR2PQ/bundle.json","state":"https://pith.science/pith/BN56XD2IMCJTN4CCJG25ZNR2PQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/BN56XD2IMCJTN4CCJG25ZNR2PQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:BN56XD2IMCJTN4CCJG25ZNR2PQ","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":"58840026ef55daca0497db17725ef912b8631c87dc2cbd3cf3e8a531d690fe70","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-03T18:33:37Z","title_canon_sha256":"802206483ed5c71bfa269c24acc1fa179f352a73220af924998525e3f89f0f76"},"schema_version":"1.0","source":{"id":"2608.02786","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.02786","created_at":"2026-08-05T00:40:02Z"},{"alias_kind":"arxiv_version","alias_value":"2608.02786v1","created_at":"2026-08-05T00:40:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.02786","created_at":"2026-08-05T00:40:02Z"},{"alias_kind":"pith_short_12","alias_value":"BN56XD2IMCJT","created_at":"2026-08-05T00:40:02Z"},{"alias_kind":"pith_short_16","alias_value":"BN56XD2IMCJTN4CC","created_at":"2026-08-05T00:40:02Z"},{"alias_kind":"pith_short_8","alias_value":"BN56XD2I","created_at":"2026-08-05T00:40:02Z"}],"graph_snapshots":[{"event_id":"sha256:b6589088bf925b9ca5f6de7ba90c9453f9dcc347e93a344a5969ff5548807245","target":"graph","created_at":"2026-08-05T00:40:02Z","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/2608.02786/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"AI systems can fail silently. The failure propagates through training loops, evaluation pipelines, and production monitoring stacks until downstream harm makes it visible. This paper introduces evaluation blindness: a measurement function M exhibits evaluation blindness with respect to failure class F when it produces readings indistinguishable from a healthy state while the system is actually failing, with no auxiliary signal flagging the gap.\n  The problem surfaces at two lifecycle stages the literature has treated separately. At training time, reward models are gamed, importance-sampling co","authors_text":"Priyanka Bajaj (Independent Researcher)","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-03T18:33:37Z","title":"Evaluation Blindness: How Silent Measurement Failures Corrupt AI Systems from Training to Deployment"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.02786","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":"sha256:94b9056de1b3c936a6578adee4cac71dbc0faaaf02dfb18c517616029f076f9e","target":"record","created_at":"2026-08-05T00:40:02Z","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":"58840026ef55daca0497db17725ef912b8631c87dc2cbd3cf3e8a531d690fe70","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-03T18:33:37Z","title_canon_sha256":"802206483ed5c71bfa269c24acc1fa179f352a73220af924998525e3f89f0f76"},"schema_version":"1.0","source":{"id":"2608.02786","kind":"arxiv","version":1}},"canonical_sha256":"0b7beb8f48609336f04249b5dcb63a7c353c98b22b11873392d84e4e53ff44d0","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0b7beb8f48609336f04249b5dcb63a7c353c98b22b11873392d84e4e53ff44d0","first_computed_at":"2026-08-05T00:40:02.220996Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-05T00:40:02.220996Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Mlo/dxXcMF+bYO0LoeBSL3/k1fSprZxldDjibuDLrppziZ1JqksRBmCLyu3pmU8kW3tpkQawaqDyKC2ok1QDBg==","signature_status":"signed_v1","signed_at":"2026-08-05T00:40:02.224004Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.02786","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:94b9056de1b3c936a6578adee4cac71dbc0faaaf02dfb18c517616029f076f9e","sha256:b6589088bf925b9ca5f6de7ba90c9453f9dcc347e93a344a5969ff5548807245"],"state_sha256":"16f72a702b4eb454a4778912d265b1a26d2631b11867263579d4d53d75f4f87d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dWBZ/7LhL2oLSxbZ+sqlIsB4gfwriChrnGPR17tadiGDR+yIGijkZ0uMnTq/5R2i4KNK12qNuW0nuTAk3LiCCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-16T05:23:52.683288Z","bundle_sha256":"9945599fe5f9e688ece565332000f95f0dbc3dcd6b438d2183dcd10a122d6c6a"}}