{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:V7IJVNPDUG3JHN5LNZCI6AXKXA","short_pith_number":"pith:V7IJVNPD","canonical_record":{"source":{"id":"2411.12709","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CY","submitted_at":"2024-11-19T18:25:30Z","cross_cats_sorted":[],"title_canon_sha256":"a01db7bbc3a9ab940dac851b00a7614eff99dbf5965ba856e0d6c4ed27104c3e","abstract_canon_sha256":"6efe0cd45ef2c5327a09553c0618e44f726b1733385a89b830774e75021f138c"},"schema_version":"1.0"},"canonical_sha256":"afd09ab5e3a1b693b7ab6e448f02eab836729099c8f71d25352a63fde1976379","source":{"kind":"arxiv","id":"2411.12709","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2411.12709","created_at":"2026-07-05T09:37:47Z"},{"alias_kind":"arxiv_version","alias_value":"2411.12709v1","created_at":"2026-07-05T09:37:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.12709","created_at":"2026-07-05T09:37:47Z"},{"alias_kind":"pith_short_12","alias_value":"V7IJVNPDUG3J","created_at":"2026-07-05T09:37:47Z"},{"alias_kind":"pith_short_16","alias_value":"V7IJVNPDUG3JHN5L","created_at":"2026-07-05T09:37:47Z"},{"alias_kind":"pith_short_8","alias_value":"V7IJVNPD","created_at":"2026-07-05T09:37:47Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:V7IJVNPDUG3JHN5LNZCI6AXKXA","target":"record","payload":{"canonical_record":{"source":{"id":"2411.12709","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CY","submitted_at":"2024-11-19T18:25:30Z","cross_cats_sorted":[],"title_canon_sha256":"a01db7bbc3a9ab940dac851b00a7614eff99dbf5965ba856e0d6c4ed27104c3e","abstract_canon_sha256":"6efe0cd45ef2c5327a09553c0618e44f726b1733385a89b830774e75021f138c"},"schema_version":"1.0"},"canonical_sha256":"afd09ab5e3a1b693b7ab6e448f02eab836729099c8f71d25352a63fde1976379","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:37:47.093189Z","signature_b64":"vSQgzBacrZKdhg295v1q3DSEVFHsjkOSGdrGxzog2nrDvhCxxdRP2vhMtJ8oQv/WjLvKpbCf0fKbg0K+L4tBDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"afd09ab5e3a1b693b7ab6e448f02eab836729099c8f71d25352a63fde1976379","last_reissued_at":"2026-07-05T09:37:47.092773Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:37:47.092773Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2411.12709","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":"2026-07-05T09:37:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"B6nY9pJwqY01JRvaY9Jq3BkABFqhafTtjAiDxu6yHEGT2bpfdcqut4jGWgHU4llKnomCPdMxFY7G/hfrLjlWAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T07:02:02.442641Z"},"content_sha256":"c5370bcba40acd152936274ee8d3b3c008a4c9e7a2df023f3b13b1e3b2f99082","schema_version":"1.0","event_id":"sha256:c5370bcba40acd152936274ee8d3b3c008a4c9e7a2df023f3b13b1e3b2f99082"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:V7IJVNPDUG3JHN5LNZCI6AXKXA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Dimensions of Generative AI Evaluation Design","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CY","authors_text":"Alexandra Chouldechova, Chad Atalla, Hanna Wallach, Jennifer Wortman Vaughan, P. Alex Dow, Solon Barocas","submitted_at":"2024-11-19T18:25:30Z","abstract_excerpt":"There are few principles or guidelines to ensure evaluations of generative AI (GenAI) models and systems are effective. To help address this gap, we propose a set of general dimensions that capture critical choices involved in GenAI evaluation design. These dimensions include the evaluation setting, the task type, the input source, the interaction style, the duration, the metric type, and the scoring method. By situating GenAI evaluations within these dimensions, we aim to guide decision-making during GenAI evaluation design and provide a structure for comparing different evaluations. We illus"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.12709","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/2411.12709/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":"2026-07-05T09:37:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HWN4sjsfbyBA31SDhxk6zvCXIYDhtDTgeUtnWD0di5D54qUWpaXYc35J2B9N75M9OApzWi9M7tZYcqjzelQsAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T07:02:02.443564Z"},"content_sha256":"7b1d38728601edca45632b37d14fbf6cb56a194d8e8160345e433e84fb17a0b9","schema_version":"1.0","event_id":"sha256:7b1d38728601edca45632b37d14fbf6cb56a194d8e8160345e433e84fb17a0b9"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/V7IJVNPDUG3JHN5LNZCI6AXKXA/bundle.json","state_url":"https://pith.science/pith/V7IJVNPDUG3JHN5LNZCI6AXKXA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/V7IJVNPDUG3JHN5LNZCI6AXKXA/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":"2026-08-15T07:02:02Z","links":{"resolver":"https://pith.science/pith/V7IJVNPDUG3JHN5LNZCI6AXKXA","bundle":"https://pith.science/pith/V7IJVNPDUG3JHN5LNZCI6AXKXA/bundle.json","state":"https://pith.science/pith/V7IJVNPDUG3JHN5LNZCI6AXKXA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/V7IJVNPDUG3JHN5LNZCI6AXKXA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:V7IJVNPDUG3JHN5LNZCI6AXKXA","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":"6efe0cd45ef2c5327a09553c0618e44f726b1733385a89b830774e75021f138c","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CY","submitted_at":"2024-11-19T18:25:30Z","title_canon_sha256":"a01db7bbc3a9ab940dac851b00a7614eff99dbf5965ba856e0d6c4ed27104c3e"},"schema_version":"1.0","source":{"id":"2411.12709","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2411.12709","created_at":"2026-07-05T09:37:47Z"},{"alias_kind":"arxiv_version","alias_value":"2411.12709v1","created_at":"2026-07-05T09:37:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.12709","created_at":"2026-07-05T09:37:47Z"},{"alias_kind":"pith_short_12","alias_value":"V7IJVNPDUG3J","created_at":"2026-07-05T09:37:47Z"},{"alias_kind":"pith_short_16","alias_value":"V7IJVNPDUG3JHN5L","created_at":"2026-07-05T09:37:47Z"},{"alias_kind":"pith_short_8","alias_value":"V7IJVNPD","created_at":"2026-07-05T09:37:47Z"}],"graph_snapshots":[{"event_id":"sha256:7b1d38728601edca45632b37d14fbf6cb56a194d8e8160345e433e84fb17a0b9","target":"graph","created_at":"2026-07-05T09:37:47Z","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/2411.12709/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"There are few principles or guidelines to ensure evaluations of generative AI (GenAI) models and systems are effective. To help address this gap, we propose a set of general dimensions that capture critical choices involved in GenAI evaluation design. These dimensions include the evaluation setting, the task type, the input source, the interaction style, the duration, the metric type, and the scoring method. By situating GenAI evaluations within these dimensions, we aim to guide decision-making during GenAI evaluation design and provide a structure for comparing different evaluations. We illus","authors_text":"Alexandra Chouldechova, Chad Atalla, Hanna Wallach, Jennifer Wortman Vaughan, P. Alex Dow, Solon Barocas","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CY","submitted_at":"2024-11-19T18:25:30Z","title":"Dimensions of Generative AI Evaluation Design"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.12709","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:c5370bcba40acd152936274ee8d3b3c008a4c9e7a2df023f3b13b1e3b2f99082","target":"record","created_at":"2026-07-05T09:37:47Z","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":"6efe0cd45ef2c5327a09553c0618e44f726b1733385a89b830774e75021f138c","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CY","submitted_at":"2024-11-19T18:25:30Z","title_canon_sha256":"a01db7bbc3a9ab940dac851b00a7614eff99dbf5965ba856e0d6c4ed27104c3e"},"schema_version":"1.0","source":{"id":"2411.12709","kind":"arxiv","version":1}},"canonical_sha256":"afd09ab5e3a1b693b7ab6e448f02eab836729099c8f71d25352a63fde1976379","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"afd09ab5e3a1b693b7ab6e448f02eab836729099c8f71d25352a63fde1976379","first_computed_at":"2026-07-05T09:37:47.092773Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:37:47.092773Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"vSQgzBacrZKdhg295v1q3DSEVFHsjkOSGdrGxzog2nrDvhCxxdRP2vhMtJ8oQv/WjLvKpbCf0fKbg0K+L4tBDw==","signature_status":"signed_v1","signed_at":"2026-07-05T09:37:47.093189Z","signed_message":"canonical_sha256_bytes"},"source_id":"2411.12709","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c5370bcba40acd152936274ee8d3b3c008a4c9e7a2df023f3b13b1e3b2f99082","sha256:7b1d38728601edca45632b37d14fbf6cb56a194d8e8160345e433e84fb17a0b9"],"state_sha256":"3610e5a703903a5cecf07532f21ba1837cf7317ba85ce0a3b62ef01591934854"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"w66bzNxY0IVCTtzA3xO6zKC7bmLhQbgmeCNY7WigLuGjX6qAxIf46T13MjYee54XWhncP8xL3GdiPvWXzbIVBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-15T07:02:02.448955Z","bundle_sha256":"285d65bd8f06f2a76e3bc13826363528ce553e286a49c8b21360636b978d2a3b"}}