{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4Y6GWTBVLMNS7HTUKCEV24KJHC","short_pith_number":"pith:4Y6GWTBV","schema_version":"1.0","canonical_sha256":"e63c6b4c355b1b2f9e7450895d714938838a68f5b861cd7e7c47b40165760929","source":{"kind":"arxiv","id":"2603.09403","version":2},"attestation_state":"computed","paper":{"title":"LLM as a Meta-Judge: Synthetic Data for NLP Evaluation Metric Validation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"David Hurych, Jind\\v{r}ich Libovick\\'y, Luk\\'a\\v{s} Eigler","submitted_at":"2026-03-10T09:15:19Z","abstract_excerpt":"Validating evaluation metrics for NLG typically relies on expensive and time-consuming human annotations, which predominantly exist only for English datasets. We propose LLM as a Meta-Judge, a scalable framework that utilizes LLMs to generate synthetic evaluation datasets via controlled semantic degradation of real data, replacing human judgment. We validate our approach using \\textit{meta-correlation}, measuring the alignment between metric rankings derived from synthetic data and those from standard human benchmarks. Experiments across Machine Translation, Question Answering, and Summarizati"},"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":"2603.09403","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-03-10T09:15:19Z","cross_cats_sorted":[],"title_canon_sha256":"d1c16b235e3e56dda7f32f96fa79f3314c59959aa655ab914ca622050f8c1a57","abstract_canon_sha256":"2bf427d896b22ae7ef4e2b08cb542df028f8acabc10c03ed0550a11ee6f58c96"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T01:03:45.279855Z","signature_b64":"/RmJrgtwrTsd9hJoUzj1aRvYfhP2z3Z8vFLIDpDeB2XFYWcGYGOYiXpalRTd21c38iUGdZl7aSIhBXhEx9zhCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e63c6b4c355b1b2f9e7450895d714938838a68f5b861cd7e7c47b40165760929","last_reissued_at":"2026-06-02T01:03:45.279330Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T01:03:45.279330Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LLM as a Meta-Judge: Synthetic Data for NLP Evaluation Metric Validation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"David Hurych, Jind\\v{r}ich Libovick\\'y, Luk\\'a\\v{s} Eigler","submitted_at":"2026-03-10T09:15:19Z","abstract_excerpt":"Validating evaluation metrics for NLG typically relies on expensive and time-consuming human annotations, which predominantly exist only for English datasets. We propose LLM as a Meta-Judge, a scalable framework that utilizes LLMs to generate synthetic evaluation datasets via controlled semantic degradation of real data, replacing human judgment. We validate our approach using \\textit{meta-correlation}, measuring the alignment between metric rankings derived from synthetic data and those from standard human benchmarks. Experiments across Machine Translation, Question Answering, and Summarizati"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.09403","kind":"arxiv","version":2},"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/2603.09403/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":"2603.09403","created_at":"2026-06-02T01:03:45.279392+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.09403v2","created_at":"2026-06-02T01:03:45.279392+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.09403","created_at":"2026-06-02T01:03:45.279392+00:00"},{"alias_kind":"pith_short_12","alias_value":"4Y6GWTBVLMNS","created_at":"2026-06-02T01:03:45.279392+00:00"},{"alias_kind":"pith_short_16","alias_value":"4Y6GWTBVLMNS7HTU","created_at":"2026-06-02T01:03:45.279392+00:00"},{"alias_kind":"pith_short_8","alias_value":"4Y6GWTBV","created_at":"2026-06-02T01:03:45.279392+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/4Y6GWTBVLMNS7HTUKCEV24KJHC","json":"https://pith.science/pith/4Y6GWTBVLMNS7HTUKCEV24KJHC.json","graph_json":"https://pith.science/api/pith-number/4Y6GWTBVLMNS7HTUKCEV24KJHC/graph.json","events_json":"https://pith.science/api/pith-number/4Y6GWTBVLMNS7HTUKCEV24KJHC/events.json","paper":"https://pith.science/paper/4Y6GWTBV"},"agent_actions":{"view_html":"https://pith.science/pith/4Y6GWTBVLMNS7HTUKCEV24KJHC","download_json":"https://pith.science/pith/4Y6GWTBVLMNS7HTUKCEV24KJHC.json","view_paper":"https://pith.science/paper/4Y6GWTBV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.09403&json=true","fetch_graph":"https://pith.science/api/pith-number/4Y6GWTBVLMNS7HTUKCEV24KJHC/graph.json","fetch_events":"https://pith.science/api/pith-number/4Y6GWTBVLMNS7HTUKCEV24KJHC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4Y6GWTBVLMNS7HTUKCEV24KJHC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4Y6GWTBVLMNS7HTUKCEV24KJHC/action/storage_attestation","attest_author":"https://pith.science/pith/4Y6GWTBVLMNS7HTUKCEV24KJHC/action/author_attestation","sign_citation":"https://pith.science/pith/4Y6GWTBVLMNS7HTUKCEV24KJHC/action/citation_signature","submit_replication":"https://pith.science/pith/4Y6GWTBVLMNS7HTUKCEV24KJHC/action/replication_record"}},"created_at":"2026-06-02T01:03:45.279392+00:00","updated_at":"2026-06-02T01:03:45.279392+00:00"}