{"record": {"aliases": ["Benchmarking and Enhancing LLMs for Rule-Intensive Review of National Standard Documents"], "artifacts": [{"id": "arxiv:2608.06312", "kind": "paper", "url": "https://arxiv.org/abs/2608.06312"}], "authors": ["Tao Wang", "Qihao Yang", "Rongjiao Liang", "Lianghong Lin", "Haitao Wang", "Xinyu Cao", "Tianyong Hao"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2608.06312", "source_url": "https://arxiv.org/abs/2608.06312"}, "categories": ["General AI", "Language & Knowledge", "cs.CL"], "description": {"en": "Evaluates LLM capability for rule-intensive review of Chinese GB/T national standard documents, including error location, dimension, and type identification across 25 diagnostic error types."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.06312", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.06312", "title": "GB/T-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["South China Normal University", "Shanghai Jiao Tong University", "China National Institute of Standardization"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2608.06312", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2608.06312", "source_url": "https://openalex.org/W7201887778"}, "key": "claire-radar:2608.06312", "modality": null, "name": "GB/T-Bench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "Medium", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "true", "origin_record_id": "bm_benchmarking-and-enhancing-llms-for-rule-i_d857dd31", "origin_source": "arxiv", "recognition_confidence": "0.6", "record_sha256": "bfc6534a7d030b35096c28d910584416b04640b6ea8635a4cf7d3211ab7e8e71", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-07T09:32:11.620138Z", "source_url": "https://arxiv.org/abs/2608.06312"}, "publisher": null, "released": "2026-08-06", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.06312", "source_url": "https://arxiv.org/abs/2608.06312"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-06312", "source": "claire_radar", "source_benchmark_id": "2608.06312", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["Benchmarking and Enhancing LLMs for Rule-Intensive Review of National Standard Documents"], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2608.06312", "observedAt": "2026-10-04T03:02:42.754940Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "state": "unavailable"}}, "status": "unavailable"}, "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "The benchmark targets a niche but high-stakes domain and includes a new multi-agent framework, which may draw attention from researchers working on document-intensive LLM applications.", "score": 55}, "capabilities": [], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-07T09:32:11.620138Z", "inputHash": "ce2548adfdfb8674a493b1eca61fa1a9c18d963d9ad7f51bdb65d3d38376cf8d", "model": "deepseek-v4-pro"}, "curation": {"canonicalNameEvidence": "we introduce GB/T-Bench, the first benchmark for the structured review of national standard documents", "canonicalNameSource": "abstract", "decisionReason": "The abstract introduces GB/T-Bench as a new benchmark with a hierarchical taxonomy, traceable error instances, and a diagnosis-oriented evaluation protocol, implying a formal scoring contract and public comparison path.", "model": "deepseek-v4-pro", "reviewedAt": "2026-09-07T09:32:11.620138Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates LLM capability for rule-intensive review of Chinese GB/T national standard documents, including error location, dimension, and type identification across 25 diagnostic error types.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "score_submission", "evidence": {"reasonCodes": ["benchmark term in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "To bridge this gap, we introduce GB/T-Bench, the first benchmark for the structured review of national standard documents."}, "familyId": "bmf_28105f8b72d7", "firstSeenAt": "2026-08-19", "heat": null, "id": "bm_benchmarking-and-enhancing-llms-for-rule-i_d857dd31", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2608.06312", "pdf": "https://arxiv.org/pdf/2608.06312", "project": null, "report": "https://arxiv.org/abs/2608.06312"}, "metrics": [], "motivation": "Large language models (LLMs) increasingly support complex professional tasks, yet their capabilities in rule-intensive document review remain insufficiently evaluated.", "name": "GB/T-Bench", "oneLine": "Evaluates LLM capability for rule-intensive review of Chinese GB/T national standard documents, including error location, dimension, and type identification across 25 diagnostic error types.", "paperTitle": "Benchmarking and Enhancing LLMs for Rule-Intensive Review of National Standard Documents", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2608.06312", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-08-19T11:05:13.395391Z"}, "publications": [], "publishers": [{"name": "GB/T-Bench Team", "organizationType": "academic-lab", "role": "benchmark-publisher", "sourceUrl": "https://arxiv.org/abs/2608.06312"}], "ranking": {"90d": {"confidence": "Low", "coverage": 0.0, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 90}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}}, "readiness": "Paper only", "recognitionConfidence": 0.6, "relation": "introduces", "releasedAt": "2026-08-06", "source": {"authors": ["Tao Wang", "Qihao Yang", "Rongjiao Liang", "Lianghong Lin", "Haitao Wang", "Xinyu Cao", "Tianyong Hao"], "categories": ["cs.CL"], "id": "2608.06312", "title": "Benchmarking and Enhancing LLMs for Rule-Intensive Review of National Standard Documents", "type": "arxiv", "url": "https://arxiv.org/abs/2608.06312"}, "sourceUpdatedAt": "2026-08-06T17:27:23Z", "topics": ["cs.CL"], "venueAttempts": [], "whyItMatters": "Provides a structured evaluation for professional document review in standards domains, addressing a gap in benchmarks focused on intrinsic quality checking of lengthy, rule-governed texts."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
