{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:FOGBG7G5ZUBHAVFHM2NN7G365A","short_pith_number":"pith:FOGBG7G5","canonical_record":{"source":{"id":"2509.19590","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-09-23T21:29:04Z","cross_cats_sorted":["cs.CY","cs.LG"],"title_canon_sha256":"43e56e86c8510864e75456102bfc21fc7730ff77217a92a32e7aae8d019c7b9e","abstract_canon_sha256":"8a616569a2c9b35e25418b6c1434b2b99e59a28b1af59d94f36a008b7b8e16b1"},"schema_version":"1.0"},"canonical_sha256":"2b8c137cddcd027054a7669adf9b7ee8054fda2bff68e5f7d3d72df2bdde0b20","source":{"kind":"arxiv","id":"2509.19590","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.19590","created_at":"2026-05-20T00:02:06Z"},{"alias_kind":"arxiv_version","alias_value":"2509.19590v2","created_at":"2026-05-20T00:02:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.19590","created_at":"2026-05-20T00:02:06Z"},{"alias_kind":"pith_short_12","alias_value":"FOGBG7G5ZUBH","created_at":"2026-05-20T00:02:06Z"},{"alias_kind":"pith_short_16","alias_value":"FOGBG7G5ZUBHAVFH","created_at":"2026-05-20T00:02:06Z"},{"alias_kind":"pith_short_8","alias_value":"FOGBG7G5","created_at":"2026-05-20T00:02:06Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:FOGBG7G5ZUBHAVFHM2NN7G365A","target":"record","payload":{"canonical_record":{"source":{"id":"2509.19590","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-09-23T21:29:04Z","cross_cats_sorted":["cs.CY","cs.LG"],"title_canon_sha256":"43e56e86c8510864e75456102bfc21fc7730ff77217a92a32e7aae8d019c7b9e","abstract_canon_sha256":"8a616569a2c9b35e25418b6c1434b2b99e59a28b1af59d94f36a008b7b8e16b1"},"schema_version":"1.0"},"canonical_sha256":"2b8c137cddcd027054a7669adf9b7ee8054fda2bff68e5f7d3d72df2bdde0b20","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:02:06.784560Z","signature_b64":"Xr/r/mEFodoP9CO9VVMtxNr4+n4TjQ+yPlRA9hxIh+0LEVZ++JvgBCD90OPId/SIQCUMbhXcuMF0M5ANAo4ADw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2b8c137cddcd027054a7669adf9b7ee8054fda2bff68e5f7d3d72df2bdde0b20","last_reissued_at":"2026-05-20T00:02:06.783851Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:02:06.783851Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2509.19590","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":"2026-05-20T00:02:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CQX5F0sRLTMRPqnMFx4BpvfKP3xd+MD9PCCIecU5BIVZChsXNnq2wcxhUxsydKJ++HKLKghXOr3MK+6dcQz2AQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-23T10:05:30.568456Z"},"content_sha256":"b70b6edc337d4790a303e04b23063c52440db73912b11ff42e9b34b41b257755","schema_version":"1.0","event_id":"sha256:b70b6edc337d4790a303e04b23063c52440db73912b11ff42e9b34b41b257755"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:FOGBG7G5ZUBHAVFHM2NN7G365A","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Position: AI Evaluations Should be Grounded on a Theory of Capability","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.CY","cs.LG"],"primary_cat":"cs.AI","authors_text":"Ashia Wilson, Nathanael Jo","submitted_at":"2025-09-23T21:29:04Z","abstract_excerpt":"Evaluations of generative models are now ubiquitous, and their outcomes critically shape public and scientific expectations of AI's capabilities. Yet skepticism about their reliability continues to grow. How can we know that a reported accuracy genuinely reflects a model's underlying performance? Although benchmark results are often presented as direct measurements of capability, in practice they are inferences: treating a score as evidence of capability already presupposes a theory of what it means to be capable at a task.\n  We argue that AI evaluations should instead be framed as inference t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.19590","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/2509.19590/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":"2026-05-20T00:02:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"D4+WaSiA9CZc7eddqJSdC+Hw6ETfammOeRkgflGvxAC2L+OuMcjU2ZvK/WCLoJgz1Cu7L1leKGj8hTBVSCgeCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-23T10:05:30.569197Z"},"content_sha256":"ebefce1a0aec35fa1922cb3cfa48c3ec2eed8889bcc33eb146a33f20dc17b3c1","schema_version":"1.0","event_id":"sha256:ebefce1a0aec35fa1922cb3cfa48c3ec2eed8889bcc33eb146a33f20dc17b3c1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FOGBG7G5ZUBHAVFHM2NN7G365A/bundle.json","state_url":"https://pith.science/pith/FOGBG7G5ZUBHAVFHM2NN7G365A/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FOGBG7G5ZUBHAVFHM2NN7G365A/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":"2026-05-23T10:05:30Z","links":{"resolver":"https://pith.science/pith/FOGBG7G5ZUBHAVFHM2NN7G365A","bundle":"https://pith.science/pith/FOGBG7G5ZUBHAVFHM2NN7G365A/bundle.json","state":"https://pith.science/pith/FOGBG7G5ZUBHAVFHM2NN7G365A/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FOGBG7G5ZUBHAVFHM2NN7G365A/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:FOGBG7G5ZUBHAVFHM2NN7G365A","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":"8a616569a2c9b35e25418b6c1434b2b99e59a28b1af59d94f36a008b7b8e16b1","cross_cats_sorted":["cs.CY","cs.LG"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-09-23T21:29:04Z","title_canon_sha256":"43e56e86c8510864e75456102bfc21fc7730ff77217a92a32e7aae8d019c7b9e"},"schema_version":"1.0","source":{"id":"2509.19590","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.19590","created_at":"2026-05-20T00:02:06Z"},{"alias_kind":"arxiv_version","alias_value":"2509.19590v2","created_at":"2026-05-20T00:02:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.19590","created_at":"2026-05-20T00:02:06Z"},{"alias_kind":"pith_short_12","alias_value":"FOGBG7G5ZUBH","created_at":"2026-05-20T00:02:06Z"},{"alias_kind":"pith_short_16","alias_value":"FOGBG7G5ZUBHAVFH","created_at":"2026-05-20T00:02:06Z"},{"alias_kind":"pith_short_8","alias_value":"FOGBG7G5","created_at":"2026-05-20T00:02:06Z"}],"graph_snapshots":[{"event_id":"sha256:ebefce1a0aec35fa1922cb3cfa48c3ec2eed8889bcc33eb146a33f20dc17b3c1","target":"graph","created_at":"2026-05-20T00:02:06Z","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/2509.19590/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Evaluations of generative models are now ubiquitous, and their outcomes critically shape public and scientific expectations of AI's capabilities. Yet skepticism about their reliability continues to grow. How can we know that a reported accuracy genuinely reflects a model's underlying performance? Although benchmark results are often presented as direct measurements of capability, in practice they are inferences: treating a score as evidence of capability already presupposes a theory of what it means to be capable at a task.\n  We argue that AI evaluations should instead be framed as inference t","authors_text":"Ashia Wilson, Nathanael Jo","cross_cats":["cs.CY","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-09-23T21:29:04Z","title":"Position: AI Evaluations Should be Grounded on a Theory of Capability"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.19590","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:b70b6edc337d4790a303e04b23063c52440db73912b11ff42e9b34b41b257755","target":"record","created_at":"2026-05-20T00:02:06Z","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":"8a616569a2c9b35e25418b6c1434b2b99e59a28b1af59d94f36a008b7b8e16b1","cross_cats_sorted":["cs.CY","cs.LG"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-09-23T21:29:04Z","title_canon_sha256":"43e56e86c8510864e75456102bfc21fc7730ff77217a92a32e7aae8d019c7b9e"},"schema_version":"1.0","source":{"id":"2509.19590","kind":"arxiv","version":2}},"canonical_sha256":"2b8c137cddcd027054a7669adf9b7ee8054fda2bff68e5f7d3d72df2bdde0b20","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2b8c137cddcd027054a7669adf9b7ee8054fda2bff68e5f7d3d72df2bdde0b20","first_computed_at":"2026-05-20T00:02:06.783851Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:02:06.783851Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Xr/r/mEFodoP9CO9VVMtxNr4+n4TjQ+yPlRA9hxIh+0LEVZ++JvgBCD90OPId/SIQCUMbhXcuMF0M5ANAo4ADw==","signature_status":"signed_v1","signed_at":"2026-05-20T00:02:06.784560Z","signed_message":"canonical_sha256_bytes"},"source_id":"2509.19590","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b70b6edc337d4790a303e04b23063c52440db73912b11ff42e9b34b41b257755","sha256:ebefce1a0aec35fa1922cb3cfa48c3ec2eed8889bcc33eb146a33f20dc17b3c1"],"state_sha256":"fdac4682083a4c8cf7226b15fc257193a5fe84b6a446b04b9d7ced3aa4764f2e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QvtG0fBJbIvslhFgqcVm2Ya4WblmfewNtnE4Z5kSDAUFIYVFxucncPXNHU2x0WdczMK6Cd1WGEwDV+TwD+6WCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-23T10:05:30.573290Z","bundle_sha256":"2a8ef96b21f2852fc5119fb2a71d1fe4c0d55234e7d17232abef5f4dfc9f2e45"}}