{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:AUKWHIXQAIQ3YATBMWJLZP5QAU","short_pith_number":"pith:AUKWHIXQ","schema_version":"1.0","canonical_sha256":"051563a2f00221bc02616592bcbfb0052c2e184b5a40810a2ba517e14a7549f5","source":{"kind":"arxiv","id":"2607.28801","version":1},"attestation_state":"computed","paper":{"title":"Benchmarks Are Not Monolithic: Sample-Level Auditing and Orchestration for LLM Evaluation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Jordan Sassoon, Philipp D. Siedler","submitted_at":"2026-07-30T19:51:55Z","abstract_excerpt":"Benchmark datasets are central to evaluating Large Language Models (LLMs), yet they are typically conceived as monolithic tasks, obscuring substantial variation in the demands of individual samples. We introduce a dataset-centric meta-evaluation framework that audits benchmark datasets at the sample level along five latent dimensions: 1. Cognitive and Knowledge Demands, 2. Language and Content Quality, 3. Task Properties, 4. Context, and 5. Ethics, Safety, and Fairness. Applying this framework, we annotate five influential benchmarks -- MMLU, ARC, WinoGrande, HellaSwag, and TruthfulQA -- revea"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2607.28801","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-30T19:51:55Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"6ea2580c60b3fa239ea8e1cb4aa1c6e255ebeba807f0d9e3c17d1b291978aca3","abstract_canon_sha256":"b401ccf4ad4c8edfa0316314a59100fe2f2e1c39bf1f414689d3d14b0d4c493f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-03T00:12:57.556921Z","signature_b64":"MpHrEQRwzo5+LJ9UYP9tZBPfuy9OzuDEQS06pCaQBq7rX+snKTTo1hVvab1hsJ8/rqKjTVvb1C0KeUx8SbgdBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"051563a2f00221bc02616592bcbfb0052c2e184b5a40810a2ba517e14a7549f5","last_reissued_at":"2026-08-03T00:12:57.554591Z","signature_status":"signed_v1","first_computed_at":"2026-08-03T00:12:57.554591Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Benchmarks Are Not Monolithic: Sample-Level Auditing and Orchestration for LLM Evaluation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Jordan Sassoon, Philipp D. Siedler","submitted_at":"2026-07-30T19:51:55Z","abstract_excerpt":"Benchmark datasets are central to evaluating Large Language Models (LLMs), yet they are typically conceived as monolithic tasks, obscuring substantial variation in the demands of individual samples. We introduce a dataset-centric meta-evaluation framework that audits benchmark datasets at the sample level along five latent dimensions: 1. Cognitive and Knowledge Demands, 2. Language and Content Quality, 3. Task Properties, 4. Context, and 5. Ethics, Safety, and Fairness. Applying this framework, we annotate five influential benchmarks -- MMLU, ARC, WinoGrande, HellaSwag, and TruthfulQA -- revea"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.28801","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/2607.28801/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2607.28801","created_at":"2026-08-03T00:12:57.556082+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.28801v1","created_at":"2026-08-03T00:12:57.556082+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.28801","created_at":"2026-08-03T00:12:57.556082+00:00"},{"alias_kind":"pith_short_12","alias_value":"AUKWHIXQAIQ3","created_at":"2026-08-03T00:12:57.556082+00:00"},{"alias_kind":"pith_short_16","alias_value":"AUKWHIXQAIQ3YATB","created_at":"2026-08-03T00:12:57.556082+00:00"},{"alias_kind":"pith_short_8","alias_value":"AUKWHIXQ","created_at":"2026-08-03T00:12:57.556082+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AUKWHIXQAIQ3YATBMWJLZP5QAU","json":"https://pith.science/pith/AUKWHIXQAIQ3YATBMWJLZP5QAU.json","graph_json":"https://pith.science/api/pith-number/AUKWHIXQAIQ3YATBMWJLZP5QAU/graph.json","events_json":"https://pith.science/api/pith-number/AUKWHIXQAIQ3YATBMWJLZP5QAU/events.json","paper":"https://pith.science/paper/AUKWHIXQ"},"agent_actions":{"view_html":"https://pith.science/pith/AUKWHIXQAIQ3YATBMWJLZP5QAU","download_json":"https://pith.science/pith/AUKWHIXQAIQ3YATBMWJLZP5QAU.json","view_paper":"https://pith.science/paper/AUKWHIXQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.28801&json=true","fetch_graph":"https://pith.science/api/pith-number/AUKWHIXQAIQ3YATBMWJLZP5QAU/graph.json","fetch_events":"https://pith.science/api/pith-number/AUKWHIXQAIQ3YATBMWJLZP5QAU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AUKWHIXQAIQ3YATBMWJLZP5QAU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AUKWHIXQAIQ3YATBMWJLZP5QAU/action/storage_attestation","attest_author":"https://pith.science/pith/AUKWHIXQAIQ3YATBMWJLZP5QAU/action/author_attestation","sign_citation":"https://pith.science/pith/AUKWHIXQAIQ3YATBMWJLZP5QAU/action/citation_signature","submit_replication":"https://pith.science/pith/AUKWHIXQAIQ3YATBMWJLZP5QAU/action/replication_record"}},"created_at":"2026-08-03T00:12:57.556082+00:00","updated_at":"2026-08-03T00:12:57.556082+00:00"}