{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:JGZAL2L3FUHAB3UUBPDWNMAARP","short_pith_number":"pith:JGZAL2L3","schema_version":"1.0","canonical_sha256":"49b205e97b2d0e00ee940bc766b0008bebd7bcb2a754dece5b442d0e71cc3df5","source":{"kind":"arxiv","id":"2409.11239","version":2},"attestation_state":"computed","paper":{"title":"LLM-as-a-Judge & Reward Model: What They Can and Cannot Do","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guijin Son, Hoyoung Lee, Hyunwoo Ko, Seunghyeok Hong, Yewon Kim","submitted_at":"2024-09-17T14:40:02Z","abstract_excerpt":"LLM-as-a-Judge and reward models are widely used alternatives of multiple-choice questions or human annotators for large language model (LLM) evaluation. Their efficacy shines in evaluating long-form responses, serving a critical role as evaluators of leaderboards and as proxies to align LLMs via reinforcement learning. However, despite their popularity, their effectiveness in diverse contexts, such as non-English prompts, factual verification, or challenging questions, remains unexplored. In this paper, we conduct a comprehensive analysis of automated evaluators, reporting several key finding"},"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":"2409.11239","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-17T14:40:02Z","cross_cats_sorted":[],"title_canon_sha256":"ee6407c24552a112819d93a93390bd5420009d9ff0d3c13e1aa3afcc8010c40f","abstract_canon_sha256":"6bd7e0ccc6931acdc636713f738c34656890f298666b57bec06d55d67e9a0d02"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:14:38.523023Z","signature_b64":"43e99WFB/ArxrMWW3UwK/bCSjWZO0uHxK4XwC1SAtALhlkgKxq3Mhcmz4xT2QT0BrGuWBqBb9v/pcCGhs8dbCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"49b205e97b2d0e00ee940bc766b0008bebd7bcb2a754dece5b442d0e71cc3df5","last_reissued_at":"2026-07-05T09:14:38.522592Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:14:38.522592Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LLM-as-a-Judge & Reward Model: What They Can and Cannot Do","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guijin Son, Hoyoung Lee, Hyunwoo Ko, Seunghyeok Hong, Yewon Kim","submitted_at":"2024-09-17T14:40:02Z","abstract_excerpt":"LLM-as-a-Judge and reward models are widely used alternatives of multiple-choice questions or human annotators for large language model (LLM) evaluation. Their efficacy shines in evaluating long-form responses, serving a critical role as evaluators of leaderboards and as proxies to align LLMs via reinforcement learning. However, despite their popularity, their effectiveness in diverse contexts, such as non-English prompts, factual verification, or challenging questions, remains unexplored. In this paper, we conduct a comprehensive analysis of automated evaluators, reporting several key finding"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.11239","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/2409.11239/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":"2409.11239","created_at":"2026-07-05T09:14:38.522655+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.11239v2","created_at":"2026-07-05T09:14:38.522655+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.11239","created_at":"2026-07-05T09:14:38.522655+00:00"},{"alias_kind":"pith_short_12","alias_value":"JGZAL2L3FUHA","created_at":"2026-07-05T09:14:38.522655+00:00"},{"alias_kind":"pith_short_16","alias_value":"JGZAL2L3FUHAB3UU","created_at":"2026-07-05T09:14:38.522655+00:00"},{"alias_kind":"pith_short_8","alias_value":"JGZAL2L3","created_at":"2026-07-05T09:14:38.522655+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":12,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26080","citing_title":"Neglected Free Lunch from Post-training: Progress Advantage for LLM Agents","ref_index":82,"is_internal_anchor":false},{"citing_arxiv_id":"2606.24004","citing_title":"Towards Spec Learning: Inference-Time Alignment from Preference Pairs","ref_index":91,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03660","citing_title":"From Answers to States: Verifiable Process-Level Evaluation of Chemical Reasoning in Large Language Models","ref_index":73,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03135","citing_title":"Uncertainty-Aware Clarification in LLM Agents with Information Gain","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.24004","citing_title":"Towards Spec Learning: Inference-Time Alignment from Preference Pairs","ref_index":91,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30479","citing_title":"COHORT: Collaborative Orchestration for Hardening via Offensive Replay on Emulated Topologies","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2411.15594","citing_title":"A Survey on LLM-as-a-Judge","ref_index":133,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16479","citing_title":"Policy-Grounded Dynamic Facet Suggestions for Job Search","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2511.14617","citing_title":"Seer: Online Context Learning for Fast Synchronous LLM Reinforcement Learning","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19782","citing_title":"KoALa-Bench: Evaluating Large Audio Language Models on Korean Speech Understanding and Faithfulness","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05371","citing_title":"LLM-as-Judge for Semantic Judging of Powerline Segmentation in UAV Inspection","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02765","citing_title":"U-Define: Designing User Workflows for Hard and Soft Constraints in LLM-Based Planning","ref_index":107,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JGZAL2L3FUHAB3UUBPDWNMAARP","json":"https://pith.science/pith/JGZAL2L3FUHAB3UUBPDWNMAARP.json","graph_json":"https://pith.science/api/pith-number/JGZAL2L3FUHAB3UUBPDWNMAARP/graph.json","events_json":"https://pith.science/api/pith-number/JGZAL2L3FUHAB3UUBPDWNMAARP/events.json","paper":"https://pith.science/paper/JGZAL2L3"},"agent_actions":{"view_html":"https://pith.science/pith/JGZAL2L3FUHAB3UUBPDWNMAARP","download_json":"https://pith.science/pith/JGZAL2L3FUHAB3UUBPDWNMAARP.json","view_paper":"https://pith.science/paper/JGZAL2L3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.11239&json=true","fetch_graph":"https://pith.science/api/pith-number/JGZAL2L3FUHAB3UUBPDWNMAARP/graph.json","fetch_events":"https://pith.science/api/pith-number/JGZAL2L3FUHAB3UUBPDWNMAARP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JGZAL2L3FUHAB3UUBPDWNMAARP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JGZAL2L3FUHAB3UUBPDWNMAARP/action/storage_attestation","attest_author":"https://pith.science/pith/JGZAL2L3FUHAB3UUBPDWNMAARP/action/author_attestation","sign_citation":"https://pith.science/pith/JGZAL2L3FUHAB3UUBPDWNMAARP/action/citation_signature","submit_replication":"https://pith.science/pith/JGZAL2L3FUHAB3UUBPDWNMAARP/action/replication_record"}},"created_at":"2026-07-05T09:14:38.522655+00:00","updated_at":"2026-07-05T09:14:38.522655+00:00"}