{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2605.26068", "kind": "paper", "url": "https://arxiv.org/abs/2605.26068"}, {"id": "gh:sufe-ailab/wsadbench", "kind": "repo", "url": "https://github.com/SUFE-AILAB/WSADBench"}], "authors": ["Xu Yao", "Siyuan Zhou", "Zhenbo Wu", "Chaochuan Hou", "Shuang Liang", "Shiping Wang", "Hailiang Huang", "Songqiao Han", "Minqi Jiang"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2605.26068", "source_url": "https://arxiv.org/abs/2605.26068"}, "categories": ["General AI", "Vision & 3D", "cs.LG"], "description": {"en": "WSADBench is a benchmark for weakly supervised anomaly detection (WSAD) that unifies evaluation across incomplete, inexact, and inaccurate supervision scenarios. It evaluates 36 algorithms across 4 modalities (tabular, video, image features, text embeddings) by systematically varying label quantity, granularity, and quality, with protocols and code provided."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2605.26068", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2605.26068", "title": "WSADBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2605.26068", "modality": null, "name": "WSADBench", "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_wsadbench_d5d448b3", "origin_source": "arxiv", "recognition_confidence": "0.95", "record_sha256": "65703ad12e911e95a8612289941ff0ecf671863ab9f4d841d61900a5dd419b4b", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2605.26068"}, "publisher": null, "released": "2026-05-25", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2605.26068", "source_url": "https://arxiv.org/abs/2605.26068"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2605-26068", "source": "claire_radar", "source_benchmark_id": "2605.26068", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Vision & 3D", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/SUFE-AILAB/WSADBench", "githubScope": "benchmark_repo", "githubStars": 8, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2605.26068", "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", "state": "unavailable"}}, "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": "694a3aefd56c44b4310cd6fb08444abd8c2be2bab95e5d467a32ca978aae6048", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "WSADBench is a benchmark for weakly supervised anomaly detection (WSAD) that unifies evaluation across incomplete, inexact, and inaccurate supervision scenarios. It evaluates 36 algorithms across 4 modalities (tabular, video, image features, text embeddings) by systematically varying label quantity, granularity, and quality, with protocols and code provided.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact named benchmark artifact released in abstract", "benchmark release stated in one sentence", "public artifact URL"], "snippet": "We release WSADBench as an open-source benchmark with code and datasets to facilitate future WSAD research: https://github.com/SUFE-AILAB/WSADBench."}, "familyId": "bmf_44fa01c8bb2a", "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/SUFE-AILAB/WSADBench", "url": "https://github.com/SUFE-AILAB/WSADBench"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/SUFE-AILAB/WSADBench", "url": "https://huggingface.co/papers/2605.26068"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/SUFE-AILAB/WSADBench.", "url": "https://arxiv.org/abs/2605.26068"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/SUFE-AILAB/WSADBench", "url": "https://arxiv.org/abs/2605.26068"}], "scope": "benchmark_repo", "url": "https://github.com/SUFE-AILAB/WSADBench", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_wsadbench_d5d448b3", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/SUFE-AILAB/WSADBench", "data": null, "hfPaper": "https://huggingface.co/papers/2605.26068", "paper": "https://arxiv.org/abs/2605.26068", "pdf": "https://arxiv.org/pdf/2605.26068", "project": null, "report": "https://arxiv.org/abs/2605.26068"}, "metrics": [], "motivation": "Weakly supervised anomaly detection (WSAD) has developed in three primary directions: incomplete, inexact, and inaccurate supervision.", "name": "WSADBench", "oneLine": "WSADBench is a benchmark for weakly supervised anomaly detection (WSAD) that unifies evaluation across incomplete, inexact, and inaccurate supervision scenarios. It evaluates 36 algorithms across 4 modalities (tabular, video, image features, text embeddings) by systematically varying label quantity, granularity, and quality, with protocols and code provided.", "paperTitle": "Rethinking Weak Supervision in Anomaly Detection: A Comprehensive Benchmark", "primaryDomain": "General AI", "publication": {"evidence": "Accepted at KDD 2026 Datasets and Benchmarks Track", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2605.26068", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "KDD 2026 Datasets and Benchmarks Track", "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "SUFE-AILAB", "organizationType": "academic-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/SUFE-AILAB/WSADBench"}], "ranking": {}, "readiness": "Runnable", "recognitionConfidence": 0.95, "relation": "introduces", "releasedAt": "2026-05-25", "source": {"authors": ["Xu Yao and Siyuan Zhou and Zhenbo Wu and Chaochuan Hou and Shuang Liang and Shiping Wang and Hailiang Huang and Songqiao Han and Minqi Jiang"], "categories": ["cs.LG", "cs.AI"], "id": "2605.26068", "title": "Rethinking Weak Supervision in Anomaly Detection: A Comprehensive Benchmark", "type": "arxiv", "url": "https://arxiv.org/abs/2605.26068"}, "sourceUpdatedAt": "2026-05-30T00:15:58Z", "topics": ["cs.LG"], "venueAttempts": [{"decisionRaw": "Accepted at KDD 2026 Datasets and Benchmarks Track", "evidence": [{"level": "author-claim", "observedAt": "2026-10-14T01:46:58.058199Z", "rawValue": "Accepted at KDD 2026 Datasets and Benchmarks Track", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2605.26068"}], "reviewStatus": "accepted", "venueName": "KDD 2026 Datasets and Benchmarks Track"}], "whyItMatters": "The field of weakly supervised anomaly detection has lacked a unified evaluation framework, with existing benchmarks isolating the three supervision types. WSADBench provides a standardized comparison that reveals performance boundaries across scenarios and informs algorithm selection for practitioners facing limited or noisy labels."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
