{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:ZSJAO2BZFL4KMOINQLMYXWCCHJ","short_pith_number":"pith:ZSJAO2BZ","schema_version":"1.0","canonical_sha256":"cc920768392af8a6390d82d98bd8423a4df3b4476640162c86f3fca468f4f1a6","source":{"kind":"arxiv","id":"2310.17631","version":2},"attestation_state":"computed","paper":{"title":"JudgeLM: Fine-tuned Large Language Models are Scalable Judges","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Lianghui Zhu, Xinggang Wang, Xinlong Wang","submitted_at":"2023-10-26T17:48:58Z","abstract_excerpt":"Evaluating Large Language Models (LLMs) in open-ended scenarios is challenging because existing benchmarks and metrics can not measure them comprehensively. To address this problem, we propose to fine-tune LLMs as scalable judges (JudgeLM) to evaluate LLMs efficiently and effectively in open-ended benchmarks. We first propose a comprehensive, large-scale, high-quality dataset containing task seeds, LLMs-generated answers, and GPT-4-generated judgments for fine-tuning high-performance judges, as well as a new benchmark for evaluating the judges. We train JudgeLM at different scales from 7B, 13B"},"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":"2310.17631","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-26T17:48:58Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8e2cf349ad0e69f1255636bd9126f6cb5ac0d40e8272f8aecbe9375c824d52b0","abstract_canon_sha256":"63ce418f30910591ef4dc186301fc1e9a4b1a7fbe035695e022f13cf3b93e90a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:21:56.379145Z","signature_b64":"zqOilBO5SjzLLeV+B7FJYBPJ8bMZ30h3ERDx+k6mhcvTdbGOr0PWaxZKQMgACa0IZsWa7fHnPDin3mI70rkACg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cc920768392af8a6390d82d98bd8423a4df3b4476640162c86f3fca468f4f1a6","last_reissued_at":"2026-07-05T10:21:56.378712Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:21:56.378712Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"JudgeLM: Fine-tuned Large Language Models are Scalable Judges","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Lianghui Zhu, Xinggang Wang, Xinlong Wang","submitted_at":"2023-10-26T17:48:58Z","abstract_excerpt":"Evaluating Large Language Models (LLMs) in open-ended scenarios is challenging because existing benchmarks and metrics can not measure them comprehensively. To address this problem, we propose to fine-tune LLMs as scalable judges (JudgeLM) to evaluate LLMs efficiently and effectively in open-ended benchmarks. We first propose a comprehensive, large-scale, high-quality dataset containing task seeds, LLMs-generated answers, and GPT-4-generated judgments for fine-tuning high-performance judges, as well as a new benchmark for evaluating the judges. We train JudgeLM at different scales from 7B, 13B"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.17631","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/2310.17631/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":"2310.17631","created_at":"2026-07-05T10:21:56.378775+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.17631v2","created_at":"2026-07-05T10:21:56.378775+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.17631","created_at":"2026-07-05T10:21:56.378775+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZSJAO2BZFL4K","created_at":"2026-07-05T10:21:56.378775+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZSJAO2BZFL4KMOIN","created_at":"2026-07-05T10:21:56.378775+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZSJAO2BZ","created_at":"2026-07-05T10:21:56.378775+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":41,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2607.08535","citing_title":"When the Judge Changes, So Does the Measurement: Auditing LLM-as-Judge Reliability","ref_index":27,"is_internal_anchor":true},{"citing_arxiv_id":"2607.07302","citing_title":"Evaluating RAG Metrics in Applied Contexts: An Experiment, Its Findings and Its Limitations","ref_index":15,"is_internal_anchor":true},{"citing_arxiv_id":"2606.24004","citing_title":"Towards Spec Learning: Inference-Time Alignment from Preference Pairs","ref_index":132,"is_internal_anchor":false},{"citing_arxiv_id":"2606.23032","citing_title":"IPO Finance Agent: Benchmark of LLM Financial Analysts Beyond Finance Agent v2, with Automated Rubric Generation, on the SpaceX (SPCX) IPO","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2606.19714","citing_title":"AURA: Adaptive Uncertainty-aware Refinement for LLM-as-a-Judge Auditing","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.19057","citing_title":"Quantifying and Auditing LLM Evaluation via Positive--Unlabeled Learning","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2606.18797","citing_title":"Beyond Scalar Scores: Exploring LLM-based Metrics for Clinical Significance Evaluation in Radiology Reports","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2606.17634","citing_title":"Prompt Perturbation for Reliable LLM Evaluation over Comparison Graphs","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2606.17162","citing_title":"MemSlides: A Hierarchical Memory Driven Agent Framework for Personalized Slide Generation with Multi-turn Local Revision","ref_index":67,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11762","citing_title":"Automated Creativity Evaluation of Language Models Across Open-Ended Tasks","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09078","citing_title":"The Hidden Bias of Process Reward Models:PRISM for Rewarding the Right Reasoning","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2606.06462","citing_title":"Benchmark Everything Everywhere All at Once","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00692","citing_title":"Self-GC: Self-Governing Context for Long-Horizon LLM Agents","ref_index":68,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03189","citing_title":"SenseJudge: Human-Centric Preference-Driven Judgment Framework","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02113","citing_title":"A Primer in Post-Training Reasoning Data: What We Know About How It Works","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01034","citing_title":"A Finite-Calibration Regime Map for LLM Judge Panels","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2606.23032","citing_title":"IPO Finance Agent: Benchmark of LLM Financial Analysts Beyond Finance Agent v2, with Automated Rubric Generation, on the SpaceX (SPCX) IPO","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2606.24004","citing_title":"Towards Spec Learning: Inference-Time Alignment from Preference Pairs","ref_index":132,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00093","citing_title":"Agreement Metrics for LLM-as-Judge Evaluation: What to Report and Why","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27761","citing_title":"AndroidDaily: A Verifiable Benchmark for Mobile GUI Agents on Real-World Closed-Source Applications","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28805","citing_title":"OmniVerifier-M1: Multimodal Meta-Verifier with Explicit Structured Recalibration","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29522","citing_title":"DeepSurvey: Enhancing Analytical Depth and Citation Reliability in Automated Survey Generation","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00497","citing_title":"\"I Strongly Suspect This Website Is a Scam\": Benchmarking PII Leakage and Detection without Defense in Autonomous Web Agents","ref_index":160,"is_internal_anchor":false},{"citing_arxiv_id":"2411.15594","citing_title":"A Survey on LLM-as-a-Judge","ref_index":229,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20809","citing_title":"Refining and Reusing Annotation Guidelines for LLM Annotation","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZSJAO2BZFL4KMOINQLMYXWCCHJ","json":"https://pith.science/pith/ZSJAO2BZFL4KMOINQLMYXWCCHJ.json","graph_json":"https://pith.science/api/pith-number/ZSJAO2BZFL4KMOINQLMYXWCCHJ/graph.json","events_json":"https://pith.science/api/pith-number/ZSJAO2BZFL4KMOINQLMYXWCCHJ/events.json","paper":"https://pith.science/paper/ZSJAO2BZ"},"agent_actions":{"view_html":"https://pith.science/pith/ZSJAO2BZFL4KMOINQLMYXWCCHJ","download_json":"https://pith.science/pith/ZSJAO2BZFL4KMOINQLMYXWCCHJ.json","view_paper":"https://pith.science/paper/ZSJAO2BZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.17631&json=true","fetch_graph":"https://pith.science/api/pith-number/ZSJAO2BZFL4KMOINQLMYXWCCHJ/graph.json","fetch_events":"https://pith.science/api/pith-number/ZSJAO2BZFL4KMOINQLMYXWCCHJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZSJAO2BZFL4KMOINQLMYXWCCHJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZSJAO2BZFL4KMOINQLMYXWCCHJ/action/storage_attestation","attest_author":"https://pith.science/pith/ZSJAO2BZFL4KMOINQLMYXWCCHJ/action/author_attestation","sign_citation":"https://pith.science/pith/ZSJAO2BZFL4KMOINQLMYXWCCHJ/action/citation_signature","submit_replication":"https://pith.science/pith/ZSJAO2BZFL4KMOINQLMYXWCCHJ/action/replication_record"}},"created_at":"2026-07-05T10:21:56.378775+00:00","updated_at":"2026-07-05T10:21:56.378775+00:00"}