{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:PUTDKNICPP6TKNP3POXIN2E7MX","short_pith_number":"pith:PUTDKNIC","schema_version":"1.0","canonical_sha256":"7d263535027bfd3535fb7bae86e89f65f8419ab05a56432d76d0d8b24b8ebf41","source":{"kind":"arxiv","id":"2608.06940","version":1},"attestation_state":"computed","paper":{"title":"Blind to the Pivotal Vote: Aggregate Independence Metrics Miss Where Verification Actually Helps","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Yang Shu","submitted_at":"2026-08-07T08:14:46Z","abstract_excerpt":"LLM judge panels are a standard evaluation tool, but prior work reports highly correlated panel errors: nine judges provide roughly the effective information of two independent ones, and aggregation closes only a small fraction of the gap. A natural remedy--a signal from a different evidence source, e.g., executing a test suite--produced no distinguishable change in the panel's effective-vote count at scale (-0.04, 95\\% CI [-0.10, +0.02]). Aggregate dependence and conditional decision utility are different questions. Elementary majority arithmetic fixes the affected set for single-ballot subst"},"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":"2608.06940","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-08-07T08:14:46Z","cross_cats_sorted":[],"title_canon_sha256":"0eed2a0e097f9d2947d65232d97b7d37cc084c131b3f91c5f21d357f42fda5c2","abstract_canon_sha256":"883e6f2d835d80ab273938014cec7980bcca20ecc33db9756c1a2f35e73e5cf6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-10T01:12:35.528783Z","signature_b64":"UcLYcuuz/thPOPnZUAfF+Kgj1l4R2hPxj94xqKhHo25vSmVtA2jgb4gfNTFCaPzBQanyyg9NObu28ej91jauCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7d263535027bfd3535fb7bae86e89f65f8419ab05a56432d76d0d8b24b8ebf41","last_reissued_at":"2026-08-10T01:12:35.526224Z","signature_status":"signed_v1","first_computed_at":"2026-08-10T01:12:35.526224Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Blind to the Pivotal Vote: Aggregate Independence Metrics Miss Where Verification Actually Helps","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Yang Shu","submitted_at":"2026-08-07T08:14:46Z","abstract_excerpt":"LLM judge panels are a standard evaluation tool, but prior work reports highly correlated panel errors: nine judges provide roughly the effective information of two independent ones, and aggregation closes only a small fraction of the gap. A natural remedy--a signal from a different evidence source, e.g., executing a test suite--produced no distinguishable change in the panel's effective-vote count at scale (-0.04, 95\\% CI [-0.10, +0.02]). Aggregate dependence and conditional decision utility are different questions. Elementary majority arithmetic fixes the affected set for single-ballot subst"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.06940","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/2608.06940/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":"2608.06940","created_at":"2026-08-10T01:12:35.527301+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.06940v1","created_at":"2026-08-10T01:12:35.527301+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.06940","created_at":"2026-08-10T01:12:35.527301+00:00"},{"alias_kind":"pith_short_12","alias_value":"PUTDKNICPP6T","created_at":"2026-08-10T01:12:35.527301+00:00"},{"alias_kind":"pith_short_16","alias_value":"PUTDKNICPP6TKNP3","created_at":"2026-08-10T01:12:35.527301+00:00"},{"alias_kind":"pith_short_8","alias_value":"PUTDKNIC","created_at":"2026-08-10T01:12:35.527301+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/PUTDKNICPP6TKNP3POXIN2E7MX","json":"https://pith.science/pith/PUTDKNICPP6TKNP3POXIN2E7MX.json","graph_json":"https://pith.science/api/pith-number/PUTDKNICPP6TKNP3POXIN2E7MX/graph.json","events_json":"https://pith.science/api/pith-number/PUTDKNICPP6TKNP3POXIN2E7MX/events.json","paper":"https://pith.science/paper/PUTDKNIC"},"agent_actions":{"view_html":"https://pith.science/pith/PUTDKNICPP6TKNP3POXIN2E7MX","download_json":"https://pith.science/pith/PUTDKNICPP6TKNP3POXIN2E7MX.json","view_paper":"https://pith.science/paper/PUTDKNIC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.06940&json=true","fetch_graph":"https://pith.science/api/pith-number/PUTDKNICPP6TKNP3POXIN2E7MX/graph.json","fetch_events":"https://pith.science/api/pith-number/PUTDKNICPP6TKNP3POXIN2E7MX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PUTDKNICPP6TKNP3POXIN2E7MX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PUTDKNICPP6TKNP3POXIN2E7MX/action/storage_attestation","attest_author":"https://pith.science/pith/PUTDKNICPP6TKNP3POXIN2E7MX/action/author_attestation","sign_citation":"https://pith.science/pith/PUTDKNICPP6TKNP3POXIN2E7MX/action/citation_signature","submit_replication":"https://pith.science/pith/PUTDKNICPP6TKNP3POXIN2E7MX/action/replication_record"}},"created_at":"2026-08-10T01:12:35.527301+00:00","updated_at":"2026-08-10T01:12:35.527301+00:00"}