{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2608.00068", "kind": "paper", "url": "https://arxiv.org/abs/2608.00068"}, {"id": "gh:safebuild/gems", "kind": "repo", "url": "https://github.com/safebuild/gems"}], "authors": ["Yi Cui", "Zilin Wang", "Yijie Xu", "Qianyi Cai", "Huizai Yao", "Shuai Jiang", "Bingzhuo Zhong", "Hui Xiong"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2608.00068", "source_url": "https://arxiv.org/abs/2608.00068"}, "categories": ["General AI", "Multimodal", "Safety", "Safety & Trustworthiness"], "description": {"en": "A benchmark for evaluating multimodal large language models on construction safety hazard identification and description, with 3,314 expert-verified task instances from over 3,000 images."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.00068", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.00068", "title": "SafeBuild-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2608.00068", "modality": null, "name": "SafeBuild-Bench", "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_safebuild-bench_538eb6ea", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "bb6feef3f2f037835a0ac0d361a9f709836832e85b81623a95febe489aed227b", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2608.00068"}, "publisher": null, "released": "2026-07-29", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.00068", "source_url": "https://arxiv.org/abs/2608.00068"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-00068", "source": "claire_radar", "source_benchmark_id": "2608.00068", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Safety & Trustworthiness", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/safebuild/gems", "githubScope": "benchmark_repo", "githubStars": 3, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 0, "hfPaperUrl": "https://huggingface.co/papers/2608.00068", "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": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}}, "status": "fresh"}, "capabilities": [], "confidence": "High", "construction": "Transform Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "8131bb22ab159a7d073ef4a03c6739f0a5bc5ebf8348ea3d6b241631f5309bb0", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "A benchmark for evaluating multimodal large language models on construction safety hazard identification and description, with 3,314 expert-verified task instances from over 3,000 images.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "evaluation protocol evidence", "public artifact URL"], "snippet": "We introduce SafeBuild-Bench, a metadata-driven benchmark for evaluating multimodal large language models on construction safety under realistic temporal and site variation."}, "familyId": "bmf_e22800c63b0a", "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/safebuild/gems", "url": "https://github.com/safebuild/gems"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/safebuild/gems", "url": "https://huggingface.co/papers/2608.00068"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/safebuild/gems.", "url": "https://arxiv.org/abs/2608.00068"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/safebuild/gems", "url": "https://arxiv.org/abs/2608.00068"}], "scope": "project_repo", "url": "https://github.com/safebuild/gems", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_safebuild-bench_538eb6ea", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/safebuild/gems", "data": null, "hfPaper": "https://huggingface.co/papers/2608.00068", "paper": "https://arxiv.org/abs/2608.00068", "pdf": "https://arxiv.org/pdf/2608.00068", "project": null, "report": "https://arxiv.org/abs/2608.00068"}, "metrics": [], "motivation": "Construction-safety models must handle concrete deployment risks, such as a worker standing near a scaffold edge without guardrails, rather than only recognize common objects in curated images.", "name": "SafeBuild-Bench", "oneLine": "A benchmark for evaluating multimodal large language models on construction safety hazard identification and description, with 3,314 expert-verified task instances from over 3,000 images.", "paperTitle": "SafeBuild-Bench: A Temporal-Robust Construction Safety Benchmark with Graph-Enhanced Data Mining", "primaryDomain": "General AI", "publication": {"evidence": "Accepted by KDD 2026. 12 pages, 6 figures", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2608.00068", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "KDD 2026", "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "SafeBuild", "organizationType": "academic-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/safebuild/gems"}], "ranking": {"90d": {"confidence": "Medium", "coverage": 0.7, "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": 0.18504537701418716, "value": 3}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": 0.0, "value": 0}}, "confidence": "Medium", "coverage": 0.7, "rank": 366, "score": 25}, "method": "attention-ranking-v13", "rank": 366, "score": 25}}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-07-29", "source": {"authors": ["Yi Cui", "Zilin Wang", "Yijie Xu", "Qianyi Cai", "Huizai Yao", "Shuai Jiang", "Bingzhuo Zhong", "and Hui Xiong"], "categories": ["cs.CV"], "id": "2608.00068", "title": "SafeBuild-Bench: A Temporal-Robust Construction Safety Benchmark with Graph-Enhanced Data Mining", "type": "arxiv", "url": "https://arxiv.org/abs/2608.00068"}, "sourceUpdatedAt": "2026-07-29T06:02:01Z", "topics": ["Multimodal", "Safety"], "venueAttempts": [{"decisionRaw": "Accepted by KDD 2026. 12 pages, 6 figures", "evidence": [{"level": "author-claim", "observedAt": "2026-10-14T01:46:58.058199Z", "rawValue": "Accepted by KDD 2026. 12 pages, 6 figures", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2608.00068"}], "reviewStatus": "accepted", "venueName": "KDD 2026"}], "whyItMatters": "Construction-safety models must handle realistic temporal and site variation; this benchmark provides a standard evaluation for hazard recognition and description, useful for deployment risk assessment."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
