{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:Y6C2V5CYAVWDDJNFS7DHQZLSMN","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":"6c6d8cdc9a91a96e54dd7e7019e29a6d11d9294c1c8bc21410f8d5a52ad293fe","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-03-20T23:35:14Z","title_canon_sha256":"8d8d32cb18b503d168f265f111f7598b03d48c127137e03d8956c95ae4504a5f"},"schema_version":"1.0","source":{"id":"2603.20562","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.20562","created_at":"2026-05-20T00:02:10Z"},{"alias_kind":"arxiv_version","alias_value":"2603.20562v3","created_at":"2026-05-20T00:02:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.20562","created_at":"2026-05-20T00:02:10Z"},{"alias_kind":"pith_short_12","alias_value":"Y6C2V5CYAVWD","created_at":"2026-05-20T00:02:10Z"},{"alias_kind":"pith_short_16","alias_value":"Y6C2V5CYAVWDDJNF","created_at":"2026-05-20T00:02:10Z"},{"alias_kind":"pith_short_8","alias_value":"Y6C2V5CY","created_at":"2026-05-20T00:02:10Z"}],"graph_snapshots":[{"event_id":"sha256:62d90e3a147835cf52718f998383878a6c73c5b9aa227fb9deb7c1151c79e80b","target":"graph","created_at":"2026-05-20T00:02:10Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"On RewardBench 2 Factuality, PCFJudge improves over direct judging by up to 7 absolute points."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That order sensitivity is a dominant, fixable source of error in listwise factuality judging and that simple aggregation over permutations does not introduce compensating biases or reduce discriminative power."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"PCFJudge improves direct LLM factuality judging by up to 7 points on RewardBench 2 Factuality by aggregating results over multiple candidate permutations."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Rerunning listwise factuality prompts over multiple candidate orderings and aggregating the results produces up to 7-point gains over direct judging."}],"snapshot_sha256":"c68f66b662db5835cf046036d05e1ddccf3636dfabb6850a40a5f541e1652e11"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"18923d2f25a092660d10d8c7010dd1c664130752d6fe7974b1e19df3d06303ab"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2603.20562/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) are now widely used as judges, yet their decisions can change under presentation choices that should be irrelevant. We study one such source of instability: candidate-order sensitivity in listwise factuality evaluation, where several answers can look similarly polished while differing substantially in hallucination risk. We introduce PCFJudge, an inference-time method that reruns the same factuality-first listwise prompt over multiple orderings of the same candidate set and aggregates the resulting scores, ranks, and uncertainty signals into a single consensus deci","authors_text":"Elsa Fan, Justin Tang, Nathan Huang, Tianyi Huang, Wenqian Chen","cross_cats":["cs.AI"],"headline":"Rerunning listwise factuality prompts over multiple candidate orderings and aggregating the results produces up to 7-point gains over direct judging.","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-03-20T23:35:14Z","title":"Permutation-Consensus Listwise Judging for Robust Factuality Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.20562","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-15T07:40:00.308621Z","id":"798d4e1b-8339-4951-a6f9-5627e1118dd8","model_set":{"reader":"grok-4.3"},"one_line_summary":"PCFJudge improves direct LLM factuality judging by up to 7 points on RewardBench 2 Factuality by aggregating results over multiple candidate permutations.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Rerunning listwise factuality prompts over multiple candidate orderings and aggregating the results produces up to 7-point gains over direct judging.","strongest_claim":"On RewardBench 2 Factuality, PCFJudge improves over direct judging by up to 7 absolute points.","weakest_assumption":"That order sensitivity is a dominant, fixable source of error in listwise factuality judging and that simple aggregation over permutations does not introduce compensating biases or reduce discriminative power."}},"verdict_id":"798d4e1b-8339-4951-a6f9-5627e1118dd8"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:cf91e03a6a28d7a2acd22968e892e99aac83d20bcddff6199a8045a38ec7901f","target":"record","created_at":"2026-05-20T00:02:10Z","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":"6c6d8cdc9a91a96e54dd7e7019e29a6d11d9294c1c8bc21410f8d5a52ad293fe","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-03-20T23:35:14Z","title_canon_sha256":"8d8d32cb18b503d168f265f111f7598b03d48c127137e03d8956c95ae4504a5f"},"schema_version":"1.0","source":{"id":"2603.20562","kind":"arxiv","version":3}},"canonical_sha256":"c785aaf458056c31a5a597c6786572636398b664b0be27ddace31cb301feb61a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c785aaf458056c31a5a597c6786572636398b664b0be27ddace31cb301feb61a","first_computed_at":"2026-05-20T00:02:10.923891Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:02:10.923891Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"gKDtxzdDF3C5WkYd68scuw/ZUZrC+6wouQ0yk2BA5diiDqpSjCfy75TnWyolhFANdBSwB2IMv8UzK7x/LG/rCQ==","signature_status":"signed_v1","signed_at":"2026-05-20T00:02:10.924658Z","signed_message":"canonical_sha256_bytes"},"source_id":"2603.20562","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:cf91e03a6a28d7a2acd22968e892e99aac83d20bcddff6199a8045a38ec7901f","sha256:62d90e3a147835cf52718f998383878a6c73c5b9aa227fb9deb7c1151c79e80b"],"state_sha256":"5a3e14188d9ad6a5aaa208645bc4a54615883b68042ba04855f940257878aa5b"}