{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3IYOCLTWTGCJ4JECL5MJK3N6LN","short_pith_number":"pith:3IYOCLTW","schema_version":"1.0","canonical_sha256":"da30e12e7699849e24825f58956dbe5b70b7bbc4ec3ed7da9cdd2bc1bb4552b3","source":{"kind":"arxiv","id":"2411.07127","version":2},"attestation_state":"computed","paper":{"title":"Benchmarking LLMs' Judgments with No Gold Standard","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Grant Schoenebeck, Shengwei Xu, Yuqing Kong, Yuxuan Lu","submitted_at":"2024-11-11T16:58:36Z","abstract_excerpt":"We introduce the GEM (Generative Estimator for Mutual Information), an evaluation metric for assessing language generation by Large Language Models (LLMs), particularly in generating informative judgments, without the need for a gold standard reference. GEM broadens the scenarios where we can benchmark LLM generation performance-from traditional ones, like machine translation and summarization, where gold standard references are readily available, to subjective tasks without clear gold standards, such as academic peer review.\n  GEM uses a generative model to estimate mutual information between"},"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":"2411.07127","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-11-11T16:58:36Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"55db84d8f3374f30a2222d76a0acdf325d48af4435d19703883580cebb001540","abstract_canon_sha256":"3c0145f9067668111c5f22e52e90203e95cb06bca1aca03b4c094b14a74909ee"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:55:25.831891Z","signature_b64":"hHIiJ4qodihHLvZj8aS4BM7G5TOYCUmWwPAfbKIUf7NYTFJV0R6XxcYHdVravAxGjo2jxyUbPlyAJjxtaSDtBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"da30e12e7699849e24825f58956dbe5b70b7bbc4ec3ed7da9cdd2bc1bb4552b3","last_reissued_at":"2026-07-05T10:55:25.831405Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:55:25.831405Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Benchmarking LLMs' Judgments with No Gold Standard","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Grant Schoenebeck, Shengwei Xu, Yuqing Kong, Yuxuan Lu","submitted_at":"2024-11-11T16:58:36Z","abstract_excerpt":"We introduce the GEM (Generative Estimator for Mutual Information), an evaluation metric for assessing language generation by Large Language Models (LLMs), particularly in generating informative judgments, without the need for a gold standard reference. GEM broadens the scenarios where we can benchmark LLM generation performance-from traditional ones, like machine translation and summarization, where gold standard references are readily available, to subjective tasks without clear gold standards, such as academic peer review.\n  GEM uses a generative model to estimate mutual information between"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.07127","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/2411.07127/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":"2411.07127","created_at":"2026-07-05T10:55:25.831460+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.07127v2","created_at":"2026-07-05T10:55:25.831460+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.07127","created_at":"2026-07-05T10:55:25.831460+00:00"},{"alias_kind":"pith_short_12","alias_value":"3IYOCLTWTGCJ","created_at":"2026-07-05T10:55:25.831460+00:00"},{"alias_kind":"pith_short_16","alias_value":"3IYOCLTWTGCJ4JEC","created_at":"2026-07-05T10:55:25.831460+00:00"},{"alias_kind":"pith_short_8","alias_value":"3IYOCLTW","created_at":"2026-07-05T10:55:25.831460+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.10810","citing_title":"Likelihood scoring for continuations of mathematical text: a self-supervised benchmark with tests for shortcut vulnerabilities","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3IYOCLTWTGCJ4JECL5MJK3N6LN","json":"https://pith.science/pith/3IYOCLTWTGCJ4JECL5MJK3N6LN.json","graph_json":"https://pith.science/api/pith-number/3IYOCLTWTGCJ4JECL5MJK3N6LN/graph.json","events_json":"https://pith.science/api/pith-number/3IYOCLTWTGCJ4JECL5MJK3N6LN/events.json","paper":"https://pith.science/paper/3IYOCLTW"},"agent_actions":{"view_html":"https://pith.science/pith/3IYOCLTWTGCJ4JECL5MJK3N6LN","download_json":"https://pith.science/pith/3IYOCLTWTGCJ4JECL5MJK3N6LN.json","view_paper":"https://pith.science/paper/3IYOCLTW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.07127&json=true","fetch_graph":"https://pith.science/api/pith-number/3IYOCLTWTGCJ4JECL5MJK3N6LN/graph.json","fetch_events":"https://pith.science/api/pith-number/3IYOCLTWTGCJ4JECL5MJK3N6LN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3IYOCLTWTGCJ4JECL5MJK3N6LN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3IYOCLTWTGCJ4JECL5MJK3N6LN/action/storage_attestation","attest_author":"https://pith.science/pith/3IYOCLTWTGCJ4JECL5MJK3N6LN/action/author_attestation","sign_citation":"https://pith.science/pith/3IYOCLTWTGCJ4JECL5MJK3N6LN/action/citation_signature","submit_replication":"https://pith.science/pith/3IYOCLTWTGCJ4JECL5MJK3N6LN/action/replication_record"}},"created_at":"2026-07-05T10:55:25.831460+00:00","updated_at":"2026-07-05T10:55:25.831460+00:00"}