{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:I5OKMWEKX2FOWWPHXVQ6ZNOACE","short_pith_number":"pith:I5OKMWEK","schema_version":"1.0","canonical_sha256":"475ca6588abe8aeb59e7bd61ecb5c0110e6ad3e9674cf3b10da68caa840aa355","source":{"kind":"arxiv","id":"2605.29800","version":1},"attestation_state":"computed","paper":{"title":"Nine Judges, Two Effective Votes: Correlated Errors Undermine LLM Evaluation Panels","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guneet Kohli","submitted_at":"2026-05-28T11:48:17Z","abstract_excerpt":"LLM-as-a-judge panels aggregate votes from multiple models, with the expectation that diverse models yield more reliable evaluations. We develop a framework to measure the true informational value of such panels and quantify how far their reliability falls short of the independent-voting ideal. Testing a panel of 9 frontier LLMs from 7 model families on three natural language inference datasets (each with 100 human annotations per item), we find that the 9 judges effectively provide only about 2 independent votes' worth of information. Roughly three-quarters of the panel's nominal independence"},"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":"2605.29800","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-28T11:48:17Z","cross_cats_sorted":[],"title_canon_sha256":"683a7f880f0d1fbc18ca14e27a13504a1aba01e00fba1b6a03b3772412fd4325","abstract_canon_sha256":"9aa37d9920370dc6dae09d061b9e4d6166a15990c66e4f423b79db51be1d9572"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T02:05:53.055083Z","signature_b64":"tN+IBbL8JgI6iSt5aKTbIRMhFAshqDVSYl8omXGLRuzy/ryPH1Z3OSH+littp8tiiKB2COE1SuvdWe/AI+QlAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"475ca6588abe8aeb59e7bd61ecb5c0110e6ad3e9674cf3b10da68caa840aa355","last_reissued_at":"2026-05-29T02:05:53.054355Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T02:05:53.054355Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Nine Judges, Two Effective Votes: Correlated Errors Undermine LLM Evaluation Panels","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guneet Kohli","submitted_at":"2026-05-28T11:48:17Z","abstract_excerpt":"LLM-as-a-judge panels aggregate votes from multiple models, with the expectation that diverse models yield more reliable evaluations. We develop a framework to measure the true informational value of such panels and quantify how far their reliability falls short of the independent-voting ideal. Testing a panel of 9 frontier LLMs from 7 model families on three natural language inference datasets (each with 100 human annotations per item), we find that the 9 judges effectively provide only about 2 independent votes' worth of information. Roughly three-quarters of the panel's nominal independence"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.29800","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/2605.29800/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":"2605.29800","created_at":"2026-05-29T02:05:53.054472+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.29800v1","created_at":"2026-05-29T02:05:53.054472+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.29800","created_at":"2026-05-29T02:05:53.054472+00:00"},{"alias_kind":"pith_short_12","alias_value":"I5OKMWEKX2FO","created_at":"2026-05-29T02:05:53.054472+00:00"},{"alias_kind":"pith_short_16","alias_value":"I5OKMWEKX2FOWWPH","created_at":"2026-05-29T02:05:53.054472+00:00"},{"alias_kind":"pith_short_8","alias_value":"I5OKMWEK","created_at":"2026-05-29T02:05:53.054472+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/I5OKMWEKX2FOWWPHXVQ6ZNOACE","json":"https://pith.science/pith/I5OKMWEKX2FOWWPHXVQ6ZNOACE.json","graph_json":"https://pith.science/api/pith-number/I5OKMWEKX2FOWWPHXVQ6ZNOACE/graph.json","events_json":"https://pith.science/api/pith-number/I5OKMWEKX2FOWWPHXVQ6ZNOACE/events.json","paper":"https://pith.science/paper/I5OKMWEK"},"agent_actions":{"view_html":"https://pith.science/pith/I5OKMWEKX2FOWWPHXVQ6ZNOACE","download_json":"https://pith.science/pith/I5OKMWEKX2FOWWPHXVQ6ZNOACE.json","view_paper":"https://pith.science/paper/I5OKMWEK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.29800&json=true","fetch_graph":"https://pith.science/api/pith-number/I5OKMWEKX2FOWWPHXVQ6ZNOACE/graph.json","fetch_events":"https://pith.science/api/pith-number/I5OKMWEKX2FOWWPHXVQ6ZNOACE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/I5OKMWEKX2FOWWPHXVQ6ZNOACE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/I5OKMWEKX2FOWWPHXVQ6ZNOACE/action/storage_attestation","attest_author":"https://pith.science/pith/I5OKMWEKX2FOWWPHXVQ6ZNOACE/action/author_attestation","sign_citation":"https://pith.science/pith/I5OKMWEKX2FOWWPHXVQ6ZNOACE/action/citation_signature","submit_replication":"https://pith.science/pith/I5OKMWEKX2FOWWPHXVQ6ZNOACE/action/replication_record"}},"created_at":"2026-05-29T02:05:53.054472+00:00","updated_at":"2026-05-29T02:05:53.054472+00:00"}