{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:FC6H2AJO56IDUM2QGSBZ6AW46K","short_pith_number":"pith:FC6H2AJO","canonical_record":{"source":{"id":"2601.23112","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CY","submitted_at":"2026-01-30T15:58:59Z","cross_cats_sorted":[],"title_canon_sha256":"6301f4ee1774215c7e265bedff330bbd39c521d18f56bf1ea0966bb322fa0db2","abstract_canon_sha256":"bafd09ee050220a350ab4d5285f3cdb75bd3a3542b2fa5aa248e7a68ff81bae0"},"schema_version":"1.0"},"canonical_sha256":"28bc7d012eef903a335034839f02dcf2864a27045a4e1ec69f5f655d0ddc8303","source":{"kind":"arxiv","id":"2601.23112","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.23112","created_at":"2026-08-06T01:38:01Z"},{"alias_kind":"arxiv_version","alias_value":"2601.23112v3","created_at":"2026-08-06T01:38:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.23112","created_at":"2026-08-06T01:38:01Z"},{"alias_kind":"pith_short_12","alias_value":"FC6H2AJO56ID","created_at":"2026-08-06T01:38:01Z"},{"alias_kind":"pith_short_16","alias_value":"FC6H2AJO56IDUM2Q","created_at":"2026-08-06T01:38:01Z"},{"alias_kind":"pith_short_8","alias_value":"FC6H2AJO","created_at":"2026-08-06T01:38:01Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:FC6H2AJO56IDUM2QGSBZ6AW46K","target":"record","payload":{"canonical_record":{"source":{"id":"2601.23112","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CY","submitted_at":"2026-01-30T15:58:59Z","cross_cats_sorted":[],"title_canon_sha256":"6301f4ee1774215c7e265bedff330bbd39c521d18f56bf1ea0966bb322fa0db2","abstract_canon_sha256":"bafd09ee050220a350ab4d5285f3cdb75bd3a3542b2fa5aa248e7a68ff81bae0"},"schema_version":"1.0"},"canonical_sha256":"28bc7d012eef903a335034839f02dcf2864a27045a4e1ec69f5f655d0ddc8303","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-06T01:38:01.674561Z","signature_b64":"2OtzbwskbsjruEsaGu0zpFPkiLiXO9S6Z7bvpV6TzQt7QjwFi0zK7pFXYDu1CpsQk2p9isfiwsyxJzWOK+wgDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"28bc7d012eef903a335034839f02dcf2864a27045a4e1ec69f5f655d0ddc8303","last_reissued_at":"2026-08-06T01:38:01.673012Z","signature_status":"signed_v1","first_computed_at":"2026-08-06T01:38:01.673012Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2601.23112","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-08-06T01:38:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"M+D8uOcYoSEIiy6vqkNbQCHJFbeO3RRcf+mMqVLeTTkHwz+jpNNB8JkOv/VVF05NkUgkvLzgU2PwrRaHrDo6BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T03:43:36.496017Z"},"content_sha256":"c113a75cdaf3e7a1852a8da35c4a04936665ae26203b877362b7b1b327295340","schema_version":"1.0","event_id":"sha256:c113a75cdaf3e7a1852a8da35c4a04936665ae26203b877362b7b1b327295340"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:FC6H2AJO56IDUM2QGSBZ6AW46K","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"How Should AI Safety Benchmarks Benchmark Safety?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CY","authors_text":"Cheng Yu, Dalia Ali, Orestis Papakyriakopoulos, Ruoxuan Cao, Severin Engelmann","submitted_at":"2026-01-30T15:58:59Z","abstract_excerpt":"AI safety benchmarks are pivotal for safety in advanced AI systems; however, they have significant technical, epistemic, and sociotechnical shortcomings. We present a review of 210 safety benchmarks that maps out common challenges in safety benchmarking, documenting failures and limitations by drawing from engineering sciences and long-established theories of risk and safety. We argue that adhering to established risk management principles, mapping the space of what can(not) be measured, developing robust probabilistic metrics, and efficiently deploying measurement theory to connect benchmarki"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.23112","kind":"arxiv","version":3},"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/2601.23112/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":"2026-08-06T01:38:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3ySovMH0pyzQswyoSsvuZ3BCPtz03TmJQ9klteKLC1M/weTI2sCRoAgfFwYz8T8sFAWcZaXmiJmzyIYV2GKZAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T03:43:36.496669Z"},"content_sha256":"4022e1a239d274a5d949127090c4cd2196bffeb33e41fb88ba4f612a6b69f048","schema_version":"1.0","event_id":"sha256:4022e1a239d274a5d949127090c4cd2196bffeb33e41fb88ba4f612a6b69f048"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FC6H2AJO56IDUM2QGSBZ6AW46K/bundle.json","state_url":"https://pith.science/pith/FC6H2AJO56IDUM2QGSBZ6AW46K/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FC6H2AJO56IDUM2QGSBZ6AW46K/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":"2026-08-08T03:43:36Z","links":{"resolver":"https://pith.science/pith/FC6H2AJO56IDUM2QGSBZ6AW46K","bundle":"https://pith.science/pith/FC6H2AJO56IDUM2QGSBZ6AW46K/bundle.json","state":"https://pith.science/pith/FC6H2AJO56IDUM2QGSBZ6AW46K/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FC6H2AJO56IDUM2QGSBZ6AW46K/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:FC6H2AJO56IDUM2QGSBZ6AW46K","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":"bafd09ee050220a350ab4d5285f3cdb75bd3a3542b2fa5aa248e7a68ff81bae0","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CY","submitted_at":"2026-01-30T15:58:59Z","title_canon_sha256":"6301f4ee1774215c7e265bedff330bbd39c521d18f56bf1ea0966bb322fa0db2"},"schema_version":"1.0","source":{"id":"2601.23112","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.23112","created_at":"2026-08-06T01:38:01Z"},{"alias_kind":"arxiv_version","alias_value":"2601.23112v3","created_at":"2026-08-06T01:38:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.23112","created_at":"2026-08-06T01:38:01Z"},{"alias_kind":"pith_short_12","alias_value":"FC6H2AJO56ID","created_at":"2026-08-06T01:38:01Z"},{"alias_kind":"pith_short_16","alias_value":"FC6H2AJO56IDUM2Q","created_at":"2026-08-06T01:38:01Z"},{"alias_kind":"pith_short_8","alias_value":"FC6H2AJO","created_at":"2026-08-06T01:38:01Z"}],"graph_snapshots":[{"event_id":"sha256:4022e1a239d274a5d949127090c4cd2196bffeb33e41fb88ba4f612a6b69f048","target":"graph","created_at":"2026-08-06T01:38:01Z","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/2601.23112/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"AI safety benchmarks are pivotal for safety in advanced AI systems; however, they have significant technical, epistemic, and sociotechnical shortcomings. We present a review of 210 safety benchmarks that maps out common challenges in safety benchmarking, documenting failures and limitations by drawing from engineering sciences and long-established theories of risk and safety. We argue that adhering to established risk management principles, mapping the space of what can(not) be measured, developing robust probabilistic metrics, and efficiently deploying measurement theory to connect benchmarki","authors_text":"Cheng Yu, Dalia Ali, Orestis Papakyriakopoulos, Ruoxuan Cao, Severin Engelmann","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CY","submitted_at":"2026-01-30T15:58:59Z","title":"How Should AI Safety Benchmarks Benchmark Safety?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.23112","kind":"arxiv","version":3},"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:c113a75cdaf3e7a1852a8da35c4a04936665ae26203b877362b7b1b327295340","target":"record","created_at":"2026-08-06T01:38:01Z","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":"bafd09ee050220a350ab4d5285f3cdb75bd3a3542b2fa5aa248e7a68ff81bae0","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CY","submitted_at":"2026-01-30T15:58:59Z","title_canon_sha256":"6301f4ee1774215c7e265bedff330bbd39c521d18f56bf1ea0966bb322fa0db2"},"schema_version":"1.0","source":{"id":"2601.23112","kind":"arxiv","version":3}},"canonical_sha256":"28bc7d012eef903a335034839f02dcf2864a27045a4e1ec69f5f655d0ddc8303","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"28bc7d012eef903a335034839f02dcf2864a27045a4e1ec69f5f655d0ddc8303","first_computed_at":"2026-08-06T01:38:01.673012Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-06T01:38:01.673012Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"2OtzbwskbsjruEsaGu0zpFPkiLiXO9S6Z7bvpV6TzQt7QjwFi0zK7pFXYDu1CpsQk2p9isfiwsyxJzWOK+wgDw==","signature_status":"signed_v1","signed_at":"2026-08-06T01:38:01.674561Z","signed_message":"canonical_sha256_bytes"},"source_id":"2601.23112","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c113a75cdaf3e7a1852a8da35c4a04936665ae26203b877362b7b1b327295340","sha256:4022e1a239d274a5d949127090c4cd2196bffeb33e41fb88ba4f612a6b69f048"],"state_sha256":"292d6ef09fb3ea3b806c840624a8cfe6477e97d8b80ec48d219962a48b9c4835"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6TCJ9W3WU57rVCKXF7CaoEaSbFTqvMOXfijH9hjTqAHLQnNcuwZ3wwELs0dK+RmZ1NKlpu7QkIYKtcXiVYCUBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T03:43:36.501128Z","bundle_sha256":"314b374b0b5bf6ec6e40e5ca19a88f9d9db542a6b9eea42ffdd5b674c8361ffe"}}