{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:FQO5XUSVPRWTC5NNWKNQ6M4LNI","short_pith_number":"pith:FQO5XUSV","canonical_record":{"source":{"id":"2407.09221","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-07-12T12:37:13Z","cross_cats_sorted":["cs.CY"],"title_canon_sha256":"763ac05d1571a9344a020beec9b4d32e8f10f2e80cd0a986474d5ce72d2cc1d8","abstract_canon_sha256":"a17f5c304ce1f30ce877b399d3afa3a99851793e8dd6cfc3300121c53403260e"},"schema_version":"1.0"},"canonical_sha256":"2c1ddbd2557c6d3175adb29b0f338b6a047da317b703aca08dccc7cb0b9a81f0","source":{"kind":"arxiv","id":"2407.09221","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.09221","created_at":"2026-07-05T08:43:16Z"},{"alias_kind":"arxiv_version","alias_value":"2407.09221v1","created_at":"2026-07-05T08:43:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.09221","created_at":"2026-07-05T08:43:16Z"},{"alias_kind":"pith_short_12","alias_value":"FQO5XUSVPRWT","created_at":"2026-07-05T08:43:16Z"},{"alias_kind":"pith_short_16","alias_value":"FQO5XUSVPRWTC5NN","created_at":"2026-07-05T08:43:16Z"},{"alias_kind":"pith_short_8","alias_value":"FQO5XUSV","created_at":"2026-07-05T08:43:16Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:FQO5XUSVPRWTC5NNWKNQ6M4LNI","target":"record","payload":{"canonical_record":{"source":{"id":"2407.09221","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-07-12T12:37:13Z","cross_cats_sorted":["cs.CY"],"title_canon_sha256":"763ac05d1571a9344a020beec9b4d32e8f10f2e80cd0a986474d5ce72d2cc1d8","abstract_canon_sha256":"a17f5c304ce1f30ce877b399d3afa3a99851793e8dd6cfc3300121c53403260e"},"schema_version":"1.0"},"canonical_sha256":"2c1ddbd2557c6d3175adb29b0f338b6a047da317b703aca08dccc7cb0b9a81f0","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:43:16.460455Z","signature_b64":"ByH5Srj55i/+6ZzUGJrtZmLB0s96iJnizmGzbemY+Rdgg0Hr36FRfhr/wghZyK7ki8T4mooX6Ie70Ev3iRFhCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2c1ddbd2557c6d3175adb29b0f338b6a047da317b703aca08dccc7cb0b9a81f0","last_reissued_at":"2026-07-05T08:43:16.459930Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:43:16.459930Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2407.09221","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":"2026-07-05T08:43:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aZ4i9DL6vHVioP0HdGcOGDcKZvl01E3swlJ9Dp0EopvR/Cz0Uf/8006em5UZnQSxMIbolyByfP66SncIZbjlDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T06:30:51.858030Z"},"content_sha256":"6a55789f85fbf7e030cfd26b73a32b8c822c400d2123f435127123704ffcfa42","schema_version":"1.0","event_id":"sha256:6a55789f85fbf7e030cfd26b73a32b8c822c400d2123f435127123704ffcfa42"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:FQO5XUSVPRWTC5NNWKNQ6M4LNI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Evaluating AI Evaluation: Perils and Prospects","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CY"],"primary_cat":"cs.AI","authors_text":"John Burden","submitted_at":"2024-07-12T12:37:13Z","abstract_excerpt":"As AI systems appear to exhibit ever-increasing capability and generality, assessing their true potential and safety becomes paramount. This paper contends that the prevalent evaluation methods for these systems are fundamentally inadequate, heightening the risks and potential hazards associated with AI. I argue that a reformation is required in the way we evaluate AI systems and that we should look towards cognitive sciences for inspiration in our approaches, which have a longstanding tradition of assessing general intelligence across diverse species. We will identify some of the difficulties"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.09221","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/2407.09221/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":"2026-07-05T08:43:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aK9xuOWjuHWZP+Pl0zdJusBrfVOYUrffUvSt2z+zwFGjkTus1bJIKmPo+Ywgzc9EEZUmydkDXvoNIpfIPN7mCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T06:30:51.858528Z"},"content_sha256":"924911a0b9a553cf8d56eac38b92d075f9a1bb740386c06114b21f6224866b19","schema_version":"1.0","event_id":"sha256:924911a0b9a553cf8d56eac38b92d075f9a1bb740386c06114b21f6224866b19"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FQO5XUSVPRWTC5NNWKNQ6M4LNI/bundle.json","state_url":"https://pith.science/pith/FQO5XUSVPRWTC5NNWKNQ6M4LNI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FQO5XUSVPRWTC5NNWKNQ6M4LNI/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":"2026-08-09T06:30:51Z","links":{"resolver":"https://pith.science/pith/FQO5XUSVPRWTC5NNWKNQ6M4LNI","bundle":"https://pith.science/pith/FQO5XUSVPRWTC5NNWKNQ6M4LNI/bundle.json","state":"https://pith.science/pith/FQO5XUSVPRWTC5NNWKNQ6M4LNI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FQO5XUSVPRWTC5NNWKNQ6M4LNI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:FQO5XUSVPRWTC5NNWKNQ6M4LNI","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":"a17f5c304ce1f30ce877b399d3afa3a99851793e8dd6cfc3300121c53403260e","cross_cats_sorted":["cs.CY"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-07-12T12:37:13Z","title_canon_sha256":"763ac05d1571a9344a020beec9b4d32e8f10f2e80cd0a986474d5ce72d2cc1d8"},"schema_version":"1.0","source":{"id":"2407.09221","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.09221","created_at":"2026-07-05T08:43:16Z"},{"alias_kind":"arxiv_version","alias_value":"2407.09221v1","created_at":"2026-07-05T08:43:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.09221","created_at":"2026-07-05T08:43:16Z"},{"alias_kind":"pith_short_12","alias_value":"FQO5XUSVPRWT","created_at":"2026-07-05T08:43:16Z"},{"alias_kind":"pith_short_16","alias_value":"FQO5XUSVPRWTC5NN","created_at":"2026-07-05T08:43:16Z"},{"alias_kind":"pith_short_8","alias_value":"FQO5XUSV","created_at":"2026-07-05T08:43:16Z"}],"graph_snapshots":[{"event_id":"sha256:924911a0b9a553cf8d56eac38b92d075f9a1bb740386c06114b21f6224866b19","target":"graph","created_at":"2026-07-05T08:43:16Z","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/2407.09221/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As AI systems appear to exhibit ever-increasing capability and generality, assessing their true potential and safety becomes paramount. This paper contends that the prevalent evaluation methods for these systems are fundamentally inadequate, heightening the risks and potential hazards associated with AI. I argue that a reformation is required in the way we evaluate AI systems and that we should look towards cognitive sciences for inspiration in our approaches, which have a longstanding tradition of assessing general intelligence across diverse species. We will identify some of the difficulties","authors_text":"John Burden","cross_cats":["cs.CY"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-07-12T12:37:13Z","title":"Evaluating AI Evaluation: Perils and Prospects"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.09221","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":"sha256:6a55789f85fbf7e030cfd26b73a32b8c822c400d2123f435127123704ffcfa42","target":"record","created_at":"2026-07-05T08:43:16Z","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":"a17f5c304ce1f30ce877b399d3afa3a99851793e8dd6cfc3300121c53403260e","cross_cats_sorted":["cs.CY"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-07-12T12:37:13Z","title_canon_sha256":"763ac05d1571a9344a020beec9b4d32e8f10f2e80cd0a986474d5ce72d2cc1d8"},"schema_version":"1.0","source":{"id":"2407.09221","kind":"arxiv","version":1}},"canonical_sha256":"2c1ddbd2557c6d3175adb29b0f338b6a047da317b703aca08dccc7cb0b9a81f0","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2c1ddbd2557c6d3175adb29b0f338b6a047da317b703aca08dccc7cb0b9a81f0","first_computed_at":"2026-07-05T08:43:16.459930Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:43:16.459930Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ByH5Srj55i/+6ZzUGJrtZmLB0s96iJnizmGzbemY+Rdgg0Hr36FRfhr/wghZyK7ki8T4mooX6Ie70Ev3iRFhCw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:43:16.460455Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.09221","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6a55789f85fbf7e030cfd26b73a32b8c822c400d2123f435127123704ffcfa42","sha256:924911a0b9a553cf8d56eac38b92d075f9a1bb740386c06114b21f6224866b19"],"state_sha256":"c5d63d3c1638de912dd904b07f19362bf184c719d5cb0f47d197a136f42a4903"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Z0tTMfIjKI9XxqwhNuV7yc6gKrQh8ujUM+6BUNEgHsN3zhirlkSvKj94sJd5jrcqUsThHoDcL+BgEb/cNYN2Aw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T06:30:51.863769Z","bundle_sha256":"94f2b2ef9c15b8d4d75f6fb0f5093bcd3878b7ade295d3d3bf11830fc6177261"}}