{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:HW2WPJIF4SNDXBWZ6KZZO7WQAL","short_pith_number":"pith:HW2WPJIF","canonical_record":{"source":{"id":"2505.10573","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CY","submitted_at":"2025-05-13T20:36:22Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"a1d203a0cb2e670bd9b64e621351ff83824777b854eddc19a6481da240a94dac","abstract_canon_sha256":"fa785fcb6a43362a953da4a55b543f514da11fa631bd2319975a37fd9a7e2968"},"schema_version":"1.0"},"canonical_sha256":"3db567a505e49a3b86d9f2b3977ed002cc86a74ca6568503a84056910a7422e4","source":{"kind":"arxiv","id":"2505.10573","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.10573","created_at":"2026-07-05T11:27:30Z"},{"alias_kind":"arxiv_version","alias_value":"2505.10573v4","created_at":"2026-07-05T11:27:30Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.10573","created_at":"2026-07-05T11:27:30Z"},{"alias_kind":"pith_short_12","alias_value":"HW2WPJIF4SND","created_at":"2026-07-05T11:27:30Z"},{"alias_kind":"pith_short_16","alias_value":"HW2WPJIF4SNDXBWZ","created_at":"2026-07-05T11:27:30Z"},{"alias_kind":"pith_short_8","alias_value":"HW2WPJIF","created_at":"2026-07-05T11:27:30Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:HW2WPJIF4SNDXBWZ6KZZO7WQAL","target":"record","payload":{"canonical_record":{"source":{"id":"2505.10573","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CY","submitted_at":"2025-05-13T20:36:22Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"a1d203a0cb2e670bd9b64e621351ff83824777b854eddc19a6481da240a94dac","abstract_canon_sha256":"fa785fcb6a43362a953da4a55b543f514da11fa631bd2319975a37fd9a7e2968"},"schema_version":"1.0"},"canonical_sha256":"3db567a505e49a3b86d9f2b3977ed002cc86a74ca6568503a84056910a7422e4","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:27:30.083031Z","signature_b64":"RcRXbk4BFG4KKw9jbQjlomO5s8tcTISAdfcBoNk3mTj/DvxqLr9FpMp9wxyU9Fa2tqusjwEbrg5sYOGJZaIYDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3db567a505e49a3b86d9f2b3977ed002cc86a74ca6568503a84056910a7422e4","last_reissued_at":"2026-07-05T11:27:30.082541Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:27:30.082541Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.10573","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:27:30Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"BJk1PVWjh94ULjEL9Xzxc5jqGNX4O3XDnJGa4YV+no3Vft4TO79BMb8rIu57H06kGIBIenRyvRxlosfii1N5Aw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T01:26:44.654361Z"},"content_sha256":"1ae3ff5aedaeb27a24ecef35d95c2f28fa0051eba8c7d6521603c3da3f6e329e","schema_version":"1.0","event_id":"sha256:1ae3ff5aedaeb27a24ecef35d95c2f28fa0051eba8c7d6521603c3da3f6e329e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:HW2WPJIF4SNDXBWZ6KZZO7WQAL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Measurement to Meaning: A Validity-Centered Framework for AI Evaluation","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CY","authors_text":"Ahmed Ahmed, Angelina Wang, Anka Reuel, Ben Domingue, Olawale Salaudeen, Sanmi Koyejo, Sudharsan Sundar, Suhana Bedi, Zachary Robertson","submitted_at":"2025-05-13T20:36:22Z","abstract_excerpt":"While the capabilities and utility of AI systems have advanced, rigorous norms for evaluating these systems have lagged. Grand claims, such as models achieving general reasoning capabilities, are supported with model performance on narrow benchmarks, like performance on graduate-level exam questions, which provide a limited and potentially misleading assessment. We provide a structured approach for reasoning about the types of evaluative claims that can be made given the available evidence. For instance, our framework helps determine whether performance on a mathematical benchmark is an indica"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.10573","kind":"arxiv","version":4},"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/2505.10573/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":"2026-07-05T11:27:30Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"DjvGpNRW8BF4nekRHJzcVU7acK2jTZscmwPo706TiS7pKY4vFu3tTIr9uExZlUwEyxWsWsU9rO/5yy1PgQP4Aw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T01:26:44.654852Z"},"content_sha256":"ae4265a5111a8ffe4a83a6dc46dead822705fb7e29972b070f9acc17a575d5af","schema_version":"1.0","event_id":"sha256:ae4265a5111a8ffe4a83a6dc46dead822705fb7e29972b070f9acc17a575d5af"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/HW2WPJIF4SNDXBWZ6KZZO7WQAL/bundle.json","state_url":"https://pith.science/pith/HW2WPJIF4SNDXBWZ6KZZO7WQAL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/HW2WPJIF4SNDXBWZ6KZZO7WQAL/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":"2026-08-07T01:26:44Z","links":{"resolver":"https://pith.science/pith/HW2WPJIF4SNDXBWZ6KZZO7WQAL","bundle":"https://pith.science/pith/HW2WPJIF4SNDXBWZ6KZZO7WQAL/bundle.json","state":"https://pith.science/pith/HW2WPJIF4SNDXBWZ6KZZO7WQAL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/HW2WPJIF4SNDXBWZ6KZZO7WQAL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:HW2WPJIF4SNDXBWZ6KZZO7WQAL","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":"fa785fcb6a43362a953da4a55b543f514da11fa631bd2319975a37fd9a7e2968","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CY","submitted_at":"2025-05-13T20:36:22Z","title_canon_sha256":"a1d203a0cb2e670bd9b64e621351ff83824777b854eddc19a6481da240a94dac"},"schema_version":"1.0","source":{"id":"2505.10573","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.10573","created_at":"2026-07-05T11:27:30Z"},{"alias_kind":"arxiv_version","alias_value":"2505.10573v4","created_at":"2026-07-05T11:27:30Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.10573","created_at":"2026-07-05T11:27:30Z"},{"alias_kind":"pith_short_12","alias_value":"HW2WPJIF4SND","created_at":"2026-07-05T11:27:30Z"},{"alias_kind":"pith_short_16","alias_value":"HW2WPJIF4SNDXBWZ","created_at":"2026-07-05T11:27:30Z"},{"alias_kind":"pith_short_8","alias_value":"HW2WPJIF","created_at":"2026-07-05T11:27:30Z"}],"graph_snapshots":[{"event_id":"sha256:ae4265a5111a8ffe4a83a6dc46dead822705fb7e29972b070f9acc17a575d5af","target":"graph","created_at":"2026-07-05T11:27:30Z","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/2505.10573/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"While the capabilities and utility of AI systems have advanced, rigorous norms for evaluating these systems have lagged. Grand claims, such as models achieving general reasoning capabilities, are supported with model performance on narrow benchmarks, like performance on graduate-level exam questions, which provide a limited and potentially misleading assessment. We provide a structured approach for reasoning about the types of evaluative claims that can be made given the available evidence. For instance, our framework helps determine whether performance on a mathematical benchmark is an indica","authors_text":"Ahmed Ahmed, Angelina Wang, Anka Reuel, Ben Domingue, Olawale Salaudeen, Sanmi Koyejo, Sudharsan Sundar, Suhana Bedi, Zachary Robertson","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CY","submitted_at":"2025-05-13T20:36:22Z","title":"Measurement to Meaning: A Validity-Centered Framework for AI Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.10573","kind":"arxiv","version":4},"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":"sha256:1ae3ff5aedaeb27a24ecef35d95c2f28fa0051eba8c7d6521603c3da3f6e329e","target":"record","created_at":"2026-07-05T11:27:30Z","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":"fa785fcb6a43362a953da4a55b543f514da11fa631bd2319975a37fd9a7e2968","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CY","submitted_at":"2025-05-13T20:36:22Z","title_canon_sha256":"a1d203a0cb2e670bd9b64e621351ff83824777b854eddc19a6481da240a94dac"},"schema_version":"1.0","source":{"id":"2505.10573","kind":"arxiv","version":4}},"canonical_sha256":"3db567a505e49a3b86d9f2b3977ed002cc86a74ca6568503a84056910a7422e4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3db567a505e49a3b86d9f2b3977ed002cc86a74ca6568503a84056910a7422e4","first_computed_at":"2026-07-05T11:27:30.082541Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:27:30.082541Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"RcRXbk4BFG4KKw9jbQjlomO5s8tcTISAdfcBoNk3mTj/DvxqLr9FpMp9wxyU9Fa2tqusjwEbrg5sYOGJZaIYDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:27:30.083031Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.10573","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1ae3ff5aedaeb27a24ecef35d95c2f28fa0051eba8c7d6521603c3da3f6e329e","sha256:ae4265a5111a8ffe4a83a6dc46dead822705fb7e29972b070f9acc17a575d5af"],"state_sha256":"3053c3f8aa84ed80b35cb4ec774f75ed269280663d7e755dd8510dbde86508b7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wD43ogsZZfgmWpYb/CHT+nMA1P5+jClylVlJm2qujGYmt7wpWJnfKHbAvaMSTMehm/BvXDf0jFpEDGQ1BmK6Cg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T01:26:44.658896Z","bundle_sha256":"079a815d378a52d864cce6f44dee7fdac23822c66522e2507d3b848b3da6e214"}}