{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2606.29920", "kind": "paper", "url": "https://arxiv.org/abs/2606.29920"}, {"id": "gh:thu-keg/ruverbench", "kind": "repo", "url": "https://github.com/THU-KEG/RuVerBench"}], "authors": ["Yangda Peng", "Yunjia Qi", "Haotian Xia", "Guanzhong He", "Xintong Shi", "Richeng Xuan", "Songyuanyi Lu", "Yixian Liu", "Zhichao Hu", "Yuhong Liu", "Hao Peng"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.29920", "source_url": "https://arxiv.org/abs/2606.29920"}, "categories": ["General AI", "Language & Knowledge", "cs.CL"], "description": {"en": "RuVerBench evaluates LLM-as-a-judge reliability for rubric verification in agentic scenarios. It includes 2,458 instances across deep research and agentic coding, each with a model-generated output, a rubric, and a human-annotated label indicating rubric satisfaction."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.29920", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.29920", "title": "RuVerBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2606.29920", "modality": null, "name": "RuVerBench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "High", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "unknown", "origin_record_id": "bm_ruverbench_289e3a74", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "c1aafef6cf9becf112c554cf0090e64184725da386b74ea398b8e5e55012e88f", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2606.29920"}, "publisher": null, "released": "2026-06-29", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.29920", "source_url": "https://arxiv.org/abs/2606.29920"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-29920", "source": "claire_radar", "source_benchmark_id": "2606.29920", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/THU-KEG/RuVerBench", "githubScope": "benchmark_repo", "githubStars": 12, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 1, "hfPaperUrl": "https://huggingface.co/papers/2606.29920", "observedAt": "2026-10-04T15:31:48.030916Z", "signalStatus": {"githubStars": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "lastSuccessfulDate": "2026-10-04", "state": "stale"}}, "status": "fresh"}, "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "20987ca85eee3fb2da929edc4f10fd99d9ae46bfb1b19dfa6a19e2dd7b656988", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "RuVerBench evaluates LLM-as-a-judge reliability for rubric verification in agentic scenarios. It includes 2,458 instances across deep research and agentic coding, each with a model-generated output, a rubric, and a human-annotated label indicating rubric satisfaction.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact named benchmark artifact released in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence", "public artifact URL"], "snippet": "We introduce RuVerBench, the first benchmark for assessing LaaJ reliability in rubric verification for agentic scenarios."}, "familyId": "bmf_646ce3fcec4f", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [{"evidence": [{"basis": "existing-code-link", "linkedUrl": "https://github.com/THU-KEG/RuVerBench", "url": "https://github.com/THU-KEG/RuVerBench"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/THU-KEG/RuVerBench", "url": "https://huggingface.co/papers/2606.29920"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/THU-KEG/RuVerBench.", "url": "https://arxiv.org/abs/2606.29920"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/THU-KEG/RuVerBench", "url": "https://arxiv.org/abs/2606.29920"}], "scope": "benchmark_repo", "url": "https://github.com/THU-KEG/RuVerBench", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_ruverbench_289e3a74", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/THU-KEG/RuVerBench", "data": null, "hfPaper": "https://huggingface.co/papers/2606.29920", "paper": "https://arxiv.org/abs/2606.29920", "pdf": "https://arxiv.org/pdf/2606.29920", "project": null, "report": "https://arxiv.org/abs/2606.29920"}, "metrics": [], "motivation": "Rubric-based scoring has become a widely used paradigm in model evaluation, typically with LLM-as-a-Judge (LaaJ) for rubric scoring.", "name": "RuVerBench", "oneLine": "RuVerBench evaluates LLM-as-a-judge reliability for rubric verification in agentic scenarios. It includes 2,458 instances across deep research and agentic coding, each with a model-generated output, a rubric, and a human-annotated label indicating rubric satisfaction.", "paperTitle": "Can LLM-as-a-Judge Reliably Verify Rubrics in Agentic Scenarios?", "primaryDomain": "General AI", "publication": {"evidence": "Accepted to EMNLP 2026", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2606.29920", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "EMNLP 2026", "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "Tsinghua University", "organizationType": "academic-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/THU-KEG/RuVerBench"}], "ranking": {}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-06-29", "source": {"authors": ["Yangda Peng", "Yunjia Qi", "Hao Peng", "Haotian Xia", "Guanzhong He", "Xintong Shi", "Richeng Xuan", "Songyuanyi Lu", "Yixian Liu", "Zhichao Hu", "Yuhong Liu", "Lei Hou", "Bin Xu", "Juanzi Li"], "categories": ["cs.CL"], "id": "2606.29920", "title": "Can LLM-as-a-Judge Reliably Verify Rubrics in Agentic Scenarios?", "type": "arxiv", "url": "https://arxiv.org/abs/2606.29920"}, "sourceUpdatedAt": "2026-06-29T07:57:23Z", "topics": ["cs.CL"], "venueAttempts": [{"decisionRaw": "Accepted to EMNLP 2026", "evidence": [{"level": "author-claim", "observedAt": "2026-10-14T01:46:58.058199Z", "rawValue": "Accepted to EMNLP 2026", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2606.29920"}], "reviewStatus": "accepted", "venueName": "EMNLP 2026"}], "whyItMatters": "Rubric-based scoring with LLM judges is common but under-validated, especially for agentic outputs. RuVerBench provides a reusable benchmark to compare judge models and strategies, enabling decisions on model selection and scoring protocol."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
