{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UN3YJ27AHVEKEPSMVQ25452RKQ","short_pith_number":"pith:UN3YJ27A","schema_version":"1.0","canonical_sha256":"a37784ebe03d48a23e4cac35de775154042f5c03cf4163c7418f1e39a1559b07","source":{"kind":"arxiv","id":"2606.13221","version":1},"attestation_state":"computed","paper":{"title":"From Uncertain Judgments to Calibrated Rankings: Conformal Elo Estimation for LLM Evaluation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bora Kargi, David Salinas","submitted_at":"2026-06-11T11:38:05Z","abstract_excerpt":"Evaluating new large language models typically requires costly human annotation campaigns at scale. LLM-as-a-judge offers a cheaper alternative, but judge scores carry systematic errors - such as position bias, self-preference, or intransitivity - that can strongly miscalibrate the resulting rankings. We quantify the resulting judge-human disagreement at two complementary levels. At the local level, we estimate per-battle uncertainty from the judge's own score differences by propagating calibrated win probabilities rather than hard labels into the Bradley-Terry procedure. This alone provides a"},"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":"2606.13221","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-11T11:38:05Z","cross_cats_sorted":[],"title_canon_sha256":"fd88d174a26d54878088aff286fbf837820b8c54d5a7b0aaafdec9dfeb31b402","abstract_canon_sha256":"4e1e6ef03c020fa27176801371726b3929619e4b5a930793c7ab739978017c96"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-12T01:09:47.296882Z","signature_b64":"oyqomYVmCPpE9ETsZwwNZ6bOLjxckMVof00mZCu9MfQs6dRcze39zUP+qPxC4Sme83BSrinENKrpgDa3TzgxBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a37784ebe03d48a23e4cac35de775154042f5c03cf4163c7418f1e39a1559b07","last_reissued_at":"2026-06-12T01:09:47.296164Z","signature_status":"signed_v1","first_computed_at":"2026-06-12T01:09:47.296164Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From Uncertain Judgments to Calibrated Rankings: Conformal Elo Estimation for LLM Evaluation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bora Kargi, David Salinas","submitted_at":"2026-06-11T11:38:05Z","abstract_excerpt":"Evaluating new large language models typically requires costly human annotation campaigns at scale. LLM-as-a-judge offers a cheaper alternative, but judge scores carry systematic errors - such as position bias, self-preference, or intransitivity - that can strongly miscalibrate the resulting rankings. We quantify the resulting judge-human disagreement at two complementary levels. At the local level, we estimate per-battle uncertainty from the judge's own score differences by propagating calibrated win probabilities rather than hard labels into the Bradley-Terry procedure. This alone provides a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.13221","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/2606.13221/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":"2606.13221","created_at":"2026-06-12T01:09:47.296273+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.13221v1","created_at":"2026-06-12T01:09:47.296273+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.13221","created_at":"2026-06-12T01:09:47.296273+00:00"},{"alias_kind":"pith_short_12","alias_value":"UN3YJ27AHVEK","created_at":"2026-06-12T01:09:47.296273+00:00"},{"alias_kind":"pith_short_16","alias_value":"UN3YJ27AHVEKEPSM","created_at":"2026-06-12T01:09:47.296273+00:00"},{"alias_kind":"pith_short_8","alias_value":"UN3YJ27A","created_at":"2026-06-12T01:09:47.296273+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/UN3YJ27AHVEKEPSMVQ25452RKQ","json":"https://pith.science/pith/UN3YJ27AHVEKEPSMVQ25452RKQ.json","graph_json":"https://pith.science/api/pith-number/UN3YJ27AHVEKEPSMVQ25452RKQ/graph.json","events_json":"https://pith.science/api/pith-number/UN3YJ27AHVEKEPSMVQ25452RKQ/events.json","paper":"https://pith.science/paper/UN3YJ27A"},"agent_actions":{"view_html":"https://pith.science/pith/UN3YJ27AHVEKEPSMVQ25452RKQ","download_json":"https://pith.science/pith/UN3YJ27AHVEKEPSMVQ25452RKQ.json","view_paper":"https://pith.science/paper/UN3YJ27A","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.13221&json=true","fetch_graph":"https://pith.science/api/pith-number/UN3YJ27AHVEKEPSMVQ25452RKQ/graph.json","fetch_events":"https://pith.science/api/pith-number/UN3YJ27AHVEKEPSMVQ25452RKQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UN3YJ27AHVEKEPSMVQ25452RKQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UN3YJ27AHVEKEPSMVQ25452RKQ/action/storage_attestation","attest_author":"https://pith.science/pith/UN3YJ27AHVEKEPSMVQ25452RKQ/action/author_attestation","sign_citation":"https://pith.science/pith/UN3YJ27AHVEKEPSMVQ25452RKQ/action/citation_signature","submit_replication":"https://pith.science/pith/UN3YJ27AHVEKEPSMVQ25452RKQ/action/replication_record"}},"created_at":"2026-06-12T01:09:47.296273+00:00","updated_at":"2026-06-12T01:09:47.296273+00:00"}