{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:CLRQDII65HV6S2W267ZN5T5CUH","short_pith_number":"pith:CLRQDII6","canonical_record":{"source":{"id":"2502.06559","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-02-10T15:25:06Z","cross_cats_sorted":[],"title_canon_sha256":"b4c7dc9b507879e966efcdbb6428e87c20f37beea026645f1649a1155e9447ab","abstract_canon_sha256":"6359c7f784125214322cbe1968992d9f40a975daac0aba5aabb0af684a570cae"},"schema_version":"1.0"},"canonical_sha256":"12e301a11ee9ebe96adaf7f2decfa2a1e27fd17637e8b12c86666bdaf233c363","source":{"kind":"arxiv","id":"2502.06559","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.06559","created_at":"2026-07-05T11:09:07Z"},{"alias_kind":"arxiv_version","alias_value":"2502.06559v2","created_at":"2026-07-05T11:09:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06559","created_at":"2026-07-05T11:09:07Z"},{"alias_kind":"pith_short_12","alias_value":"CLRQDII65HV6","created_at":"2026-07-05T11:09:07Z"},{"alias_kind":"pith_short_16","alias_value":"CLRQDII65HV6S2W2","created_at":"2026-07-05T11:09:07Z"},{"alias_kind":"pith_short_8","alias_value":"CLRQDII6","created_at":"2026-07-05T11:09:07Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:CLRQDII65HV6S2W267ZN5T5CUH","target":"record","payload":{"canonical_record":{"source":{"id":"2502.06559","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-02-10T15:25:06Z","cross_cats_sorted":[],"title_canon_sha256":"b4c7dc9b507879e966efcdbb6428e87c20f37beea026645f1649a1155e9447ab","abstract_canon_sha256":"6359c7f784125214322cbe1968992d9f40a975daac0aba5aabb0af684a570cae"},"schema_version":"1.0"},"canonical_sha256":"12e301a11ee9ebe96adaf7f2decfa2a1e27fd17637e8b12c86666bdaf233c363","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:09:07.929024Z","signature_b64":"cJNs5F5fW5hobBrPI7YP7FJuNlF7N7v2T81Xp7Wvc/TgYED2/3NBI3Y0PLJD+kwt+4H6Wvkr3raWGO0YSnNjCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"12e301a11ee9ebe96adaf7f2decfa2a1e27fd17637e8b12c86666bdaf233c363","last_reissued_at":"2026-07-05T11:09:07.928541Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:09:07.928541Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.06559","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:09:07Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"1zJayauOM4nV1syjAKCkf0ybEsLfqCyW9hSV5T+ofUijQrzu2bwVxpYcoMpuvYQDXLcdq6rakhiZH4K8D5jmAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T12:32:44.170575Z"},"content_sha256":"5813ce1568a362e413d605aa7cacf66e98a88d206d03f3385765ff09c896c079","schema_version":"1.0","event_id":"sha256:5813ce1568a362e413d605aa7cacf66e98a88d206d03f3385765ff09c896c079"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:CLRQDII65HV6S2W267ZN5T5CUH","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Can We Trust AI Benchmarks? An Interdisciplinary Review of Current Issues in AI Evaluation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Arman Noroozian, David Fernandez-Llorca, Emilia Gomez, Erasmo Purificato, Guillaume Chaslot, Joao Vinagre, Maria Eriksson","submitted_at":"2025-02-10T15:25:06Z","abstract_excerpt":"Quantitative Artificial Intelligence (AI) Benchmarks have emerged as fundamental tools for evaluating the performance, capability, and safety of AI models and systems. Currently, they shape the direction of AI development and are playing an increasingly prominent role in regulatory frameworks. As their influence grows, however, so too does concerns about how and with what effects they evaluate highly sensitive topics such as capabilities, including high-impact capabilities, safety and systemic risks. This paper presents an interdisciplinary meta-review of about 100 studies that discuss shortco"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06559","kind":"arxiv","version":2},"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/2502.06559/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":"2026-07-05T11:09:07Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EfkJE5H1UUeD4/orjF7TdCNwgm4bEBIqNJRby6lFsQR37HKWT2LsvJsE96HKWorj1dwzUIiQ70wKAf0ni4VUDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T12:32:44.170961Z"},"content_sha256":"594de43776bdba58a527b6928ef9f0c4b1508986b5d0b7da609c3bd5554d3f7c","schema_version":"1.0","event_id":"sha256:594de43776bdba58a527b6928ef9f0c4b1508986b5d0b7da609c3bd5554d3f7c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/CLRQDII65HV6S2W267ZN5T5CUH/bundle.json","state_url":"https://pith.science/pith/CLRQDII65HV6S2W267ZN5T5CUH/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/CLRQDII65HV6S2W267ZN5T5CUH/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":"2026-07-07T12:32:44Z","links":{"resolver":"https://pith.science/pith/CLRQDII65HV6S2W267ZN5T5CUH","bundle":"https://pith.science/pith/CLRQDII65HV6S2W267ZN5T5CUH/bundle.json","state":"https://pith.science/pith/CLRQDII65HV6S2W267ZN5T5CUH/state.json","well_known_bundle":"https://pith.science/.well-known/pith/CLRQDII65HV6S2W267ZN5T5CUH/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:CLRQDII65HV6S2W267ZN5T5CUH","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":"6359c7f784125214322cbe1968992d9f40a975daac0aba5aabb0af684a570cae","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-02-10T15:25:06Z","title_canon_sha256":"b4c7dc9b507879e966efcdbb6428e87c20f37beea026645f1649a1155e9447ab"},"schema_version":"1.0","source":{"id":"2502.06559","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.06559","created_at":"2026-07-05T11:09:07Z"},{"alias_kind":"arxiv_version","alias_value":"2502.06559v2","created_at":"2026-07-05T11:09:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06559","created_at":"2026-07-05T11:09:07Z"},{"alias_kind":"pith_short_12","alias_value":"CLRQDII65HV6","created_at":"2026-07-05T11:09:07Z"},{"alias_kind":"pith_short_16","alias_value":"CLRQDII65HV6S2W2","created_at":"2026-07-05T11:09:07Z"},{"alias_kind":"pith_short_8","alias_value":"CLRQDII6","created_at":"2026-07-05T11:09:07Z"}],"graph_snapshots":[{"event_id":"sha256:594de43776bdba58a527b6928ef9f0c4b1508986b5d0b7da609c3bd5554d3f7c","target":"graph","created_at":"2026-07-05T11:09:07Z","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/2502.06559/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Quantitative Artificial Intelligence (AI) Benchmarks have emerged as fundamental tools for evaluating the performance, capability, and safety of AI models and systems. Currently, they shape the direction of AI development and are playing an increasingly prominent role in regulatory frameworks. As their influence grows, however, so too does concerns about how and with what effects they evaluate highly sensitive topics such as capabilities, including high-impact capabilities, safety and systemic risks. This paper presents an interdisciplinary meta-review of about 100 studies that discuss shortco","authors_text":"Arman Noroozian, David Fernandez-Llorca, Emilia Gomez, Erasmo Purificato, Guillaume Chaslot, Joao Vinagre, Maria Eriksson","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-02-10T15:25:06Z","title":"Can We Trust AI Benchmarks? An Interdisciplinary Review of Current Issues in AI Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06559","kind":"arxiv","version":2},"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":"sha256:5813ce1568a362e413d605aa7cacf66e98a88d206d03f3385765ff09c896c079","target":"record","created_at":"2026-07-05T11:09:07Z","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":"6359c7f784125214322cbe1968992d9f40a975daac0aba5aabb0af684a570cae","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-02-10T15:25:06Z","title_canon_sha256":"b4c7dc9b507879e966efcdbb6428e87c20f37beea026645f1649a1155e9447ab"},"schema_version":"1.0","source":{"id":"2502.06559","kind":"arxiv","version":2}},"canonical_sha256":"12e301a11ee9ebe96adaf7f2decfa2a1e27fd17637e8b12c86666bdaf233c363","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"12e301a11ee9ebe96adaf7f2decfa2a1e27fd17637e8b12c86666bdaf233c363","first_computed_at":"2026-07-05T11:09:07.928541Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:09:07.928541Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"cJNs5F5fW5hobBrPI7YP7FJuNlF7N7v2T81Xp7Wvc/TgYED2/3NBI3Y0PLJD+kwt+4H6Wvkr3raWGO0YSnNjCw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:09:07.929024Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.06559","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5813ce1568a362e413d605aa7cacf66e98a88d206d03f3385765ff09c896c079","sha256:594de43776bdba58a527b6928ef9f0c4b1508986b5d0b7da609c3bd5554d3f7c"],"state_sha256":"aa0703f546cf7b41d13f07e48e5e392cabdeec2b177f0b6b7c6cb4507406859a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Y4z3k7C0tuHND6bBeCglTllqea79MpK9vySkFxt0DjqI+VOvPA1pSi6Q5FwnQN+x9aBR6+LYOXORp5liDjZ0Cw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T12:32:44.172928Z","bundle_sha256":"53a1477a8a576e26d7588b5632c71a8b9778a5f26cff45b78ed9825c193e96dc"}}