{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:FKCMVDEBTMMBWDEOLOWC3SO5VT","short_pith_number":"pith:FKCMVDEB","schema_version":"1.0","canonical_sha256":"2a84ca8c819b181b0c8e5bac2dc9ddacccc62c4628d9bbd13c99fab2cc4de938","source":{"kind":"arxiv","id":"2607.13707","version":1},"attestation_state":"computed","paper":{"title":"The Test Oracle Problem in Synthetic LLM-as-Judge Corpora: Disappearance, Distortion and a Validation Protocol","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Burdur, Mehmet Akif Ersoy University, Serkan Ball{\\i} (Department of Software Engineering, T\\\"urkiye)","submitted_at":"2026-07-15T11:12:54Z","abstract_excerpt":"Studies of bias in LLM-as-judge systems typically build synthetic corpora by prompting an LLM to generate a hallucinated answer to pair with a factual one, then presenting both to a judge. We report a case in which this generation step silently failed, and use it to argue that the failure mode is structural rather than incidental. In a multilingual (Turkish/English) faithfulness-judgment corpus, a decoding-budget parameter shared between judging and generation calls truncated one producer's hallucinated answers to a few words. The resulting items produced a large, statistically robust effect: "},"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":"2607.13707","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-07-15T11:12:54Z","cross_cats_sorted":[],"title_canon_sha256":"5283c5ff420961af83bd662623943b86ce9224783074db212b9d9d959e62022e","abstract_canon_sha256":"7423ab29a5603da705c6494dfec13a138cfed023a51e8a5bc75f34e57379b804"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-16T01:23:02.255307Z","signature_b64":"qGA7UOg3r1U6LpUeIdW5PJQHYZHLsgyb7JZ90DGMLusuxcNQFB7J9KkioHIR/jWVkGt2Gs2nDPDghhGb+KglAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2a84ca8c819b181b0c8e5bac2dc9ddacccc62c4628d9bbd13c99fab2cc4de938","last_reissued_at":"2026-07-16T01:23:02.254118Z","signature_status":"signed_v1","first_computed_at":"2026-07-16T01:23:02.254118Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Test Oracle Problem in Synthetic LLM-as-Judge Corpora: Disappearance, Distortion and a Validation Protocol","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Burdur, Mehmet Akif Ersoy University, Serkan Ball{\\i} (Department of Software Engineering, T\\\"urkiye)","submitted_at":"2026-07-15T11:12:54Z","abstract_excerpt":"Studies of bias in LLM-as-judge systems typically build synthetic corpora by prompting an LLM to generate a hallucinated answer to pair with a factual one, then presenting both to a judge. We report a case in which this generation step silently failed, and use it to argue that the failure mode is structural rather than incidental. In a multilingual (Turkish/English) faithfulness-judgment corpus, a decoding-budget parameter shared between judging and generation calls truncated one producer's hallucinated answers to a few words. The resulting items produced a large, statistically robust effect: "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.13707","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/2607.13707/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":"2607.13707","created_at":"2026-07-16T01:23:02.254860+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.13707v1","created_at":"2026-07-16T01:23:02.254860+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.13707","created_at":"2026-07-16T01:23:02.254860+00:00"},{"alias_kind":"pith_short_12","alias_value":"FKCMVDEBTMMB","created_at":"2026-07-16T01:23:02.254860+00:00"},{"alias_kind":"pith_short_16","alias_value":"FKCMVDEBTMMBWDEO","created_at":"2026-07-16T01:23:02.254860+00:00"},{"alias_kind":"pith_short_8","alias_value":"FKCMVDEB","created_at":"2026-07-16T01:23:02.254860+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/FKCMVDEBTMMBWDEOLOWC3SO5VT","json":"https://pith.science/pith/FKCMVDEBTMMBWDEOLOWC3SO5VT.json","graph_json":"https://pith.science/api/pith-number/FKCMVDEBTMMBWDEOLOWC3SO5VT/graph.json","events_json":"https://pith.science/api/pith-number/FKCMVDEBTMMBWDEOLOWC3SO5VT/events.json","paper":"https://pith.science/paper/FKCMVDEB"},"agent_actions":{"view_html":"https://pith.science/pith/FKCMVDEBTMMBWDEOLOWC3SO5VT","download_json":"https://pith.science/pith/FKCMVDEBTMMBWDEOLOWC3SO5VT.json","view_paper":"https://pith.science/paper/FKCMVDEB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.13707&json=true","fetch_graph":"https://pith.science/api/pith-number/FKCMVDEBTMMBWDEOLOWC3SO5VT/graph.json","fetch_events":"https://pith.science/api/pith-number/FKCMVDEBTMMBWDEOLOWC3SO5VT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FKCMVDEBTMMBWDEOLOWC3SO5VT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FKCMVDEBTMMBWDEOLOWC3SO5VT/action/storage_attestation","attest_author":"https://pith.science/pith/FKCMVDEBTMMBWDEOLOWC3SO5VT/action/author_attestation","sign_citation":"https://pith.science/pith/FKCMVDEBTMMBWDEOLOWC3SO5VT/action/citation_signature","submit_replication":"https://pith.science/pith/FKCMVDEBTMMBWDEOLOWC3SO5VT/action/replication_record"}},"created_at":"2026-07-16T01:23:02.254860+00:00","updated_at":"2026-07-16T01:23:02.254860+00:00"}