{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2608.17607", "kind": "paper", "url": "https://arxiv.org/abs/2608.17607"}, {"id": "hf:liubw/pathoargus-bench", "kind": "dataset", "url": "https://huggingface.co/datasets/liubw/PathoArgus-Bench"}], "authors": ["Bowen Liu", "Qixiang Zhang", "Xiaomeng Li"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2608.17607", "source_url": "https://arxiv.org/abs/2608.17607"}, "categories": ["Biology & Drug Discovery", "Evidence grounding", "Long Context", "Long-context reasoning", "Medical image reasoning", "Reasoning", "Vision & 3D"], "description": {"en": "PathoArgus-Bench evaluates evidence-grounded visual reasoning in whole-slide pathology. It comprises 22,078 multiple-choice questions from 4,913 patients across 15 TCGA projects, testing availability, accessibility, use, and responsiveness of evidence under a fixed reader budget."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.17607", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.17607", "title": "PathoArgus-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Hong Kong University of Science and Technology"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2608.17607", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2608.17607", "source_url": "https://openalex.org/W7203743890"}, "key": "claire-radar:2608.17607", "modality": null, "name": "PathoArgus-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_pathoargus-bench_2bed53e5", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "6afaca313f0c082d2ad94ed1c634c2a79d920652eaa2dc771b0a1aff9d11d1f8", "relation": "introduces", "review_state": "source-reviewed", "reviewed_at": "2026-08-20", "source_url": "https://arxiv.org/abs/2608.17607"}, "publisher": null, "released": "2026-08-18", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.17607", "source_url": "https://arxiv.org/abs/2608.17607"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-17607", "source": "claire_radar", "source_benchmark_id": "2608.17607", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["Biology & Drug Discovery"], "area": "Vision & 3D", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": "liubw/PathoArgus-Bench", "hfDatasetDownloads": 37, "hfDatasetLikes": 0, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2608.17607", "observedAt": "2026-10-04T03:02:42.754940Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"checkedAt": "2026-10-04T15:31:48.030916Z", "lastSuccessfulDate": "2026-09-30", "state": "stale"}, "hfDatasetLikes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "lastSuccessfulDate": "2026-09-30", "state": "stale"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "state": "unavailable"}}, "status": "stale"}, "availability": {"evaluatorStatus": "not_found", "githubStatus": "not_found", "hfDatasetStatus": "available", "paperStatus": "available", "submissionStatus": "not_found"}, "capabilities": ["Medical image reasoning", "Long-context reasoning", "Evidence grounding"], "confidence": "High", "construction": "Unknown", "constructionDetail": "PathoArgus-Bench measures case-level pathology reasoning over whole-slide images, including diagnostic findings, staging and exact evidence grounding.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "459e70fb59ee84661c78fd4108f27802d708370fbf5061622b60b457c6dfdb2f", "model": "deepseek-v4-flash"}, "curation": {"reviewedAt": "2026-08-20", "sources": ["https://arxiv.org/abs/2608.17607", "https://arxiv.org/html/2608.17607", "https://huggingface.co/datasets/liubw/PathoArgus-Bench"], "state": "source-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "PathoArgus-Bench evaluates evidence-grounded visual reasoning in whole-slide pathology. It comprises 22,078 multiple-choice questions from 4,913 patients across 15 TCGA projects, testing availability, accessibility, use, and responsiveness of evidence under a fixed reader budget.", "detail": {"protocol": {"primaryMetric": "Overall accuracy and QExact", "tasks": "22,078 four-choice questions from 4,913 patients and 5,400 whole-slide images", "version": "arXiv v1"}, "taskBreakdown": ["Anatomic site", "Histologic type", "Tumor grade", "Mitotic activity", "Local invasion", "Surgical margins", "Nodal status", "Pathologic staging", "Cross-slide integration", "Evidence grounding"]}, "domainCuration": {"method": "primary-source-review-v1", "state": "reviewed"}, "evaluationMode": "score_submission", "evidence": {"reasonCodes": ["exact named benchmark artifact released in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We introduce PathoArgus-Bench, a benchmark and evaluation protocol that explicitly tests the full evidence chain: availability, accessibility, use, and responsiveness."}, "familyId": "bmf_9d5dfa0f7014", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_pathoargus-bench_2bed53e5", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": ["Pharma & Biotech"], "links": {"code": null, "data": "https://huggingface.co/datasets/liubw/PathoArgus-Bench", "hfPaper": "https://huggingface.co/papers/2608.17607", "paper": "https://arxiv.org/abs/2608.17607", "pdf": "https://arxiv.org/pdf/2608.17607", "project": null, "report": "https://arxiv.org/abs/2608.17607"}, "metrics": [], "motivation": "Whole-slide pathology reasoning requires models to integrate gigapixel-scale visual evidence across complete case-linked slides, yet current question-answering benchmarks primarily measure final answer accuracy--a metric vulnerable to linguistic priors and benchmark regularities, and insufficient to establish that predictions are grounded in the supplied tissue.", "name": "PathoArgus-Bench", "oneLine": "Evaluates evidence-grounded visual reasoning over complete gigapixel whole-slide and multi-slide pathology cases.", "paperTitle": "PathoArgus: Advancing Evidence-Grounded Long-Context Visual Reasoning across Gigapixel Whole-Slide and Multi-Slide Case Contexts", "primaryDomain": "Biology & Drug Discovery", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2608.17607", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "PathoArgus-Bench Team", "organizationType": "academic-lab", "role": "benchmark-publisher", "sourceUrl": "https://huggingface.co/datasets/liubw/PathoArgus-Bench"}], "ranking": {"90d": {"confidence": "Low", "coverage": 0.3, "datasetDownloadRank": 82, "datasetRankPopulation": 85, "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": 0.37459834221054866, "value": 37}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.3, "rank": 95, "score": 46}, "method": "attention-ranking-v13", "rank": 95, "score": 46}}, "readiness": "Inspectable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-08-18", "source": {"authors": ["Bowen Liu", "Qixiang Zhang", "Xiaomeng Li"], "categories": ["cs.CV"], "id": "2608.17607", "title": "PathoArgus: Advancing Evidence-Grounded Long-Context Visual Reasoning across Gigapixel Whole-Slide and Multi-Slide Case Contexts", "type": "arxiv", "url": "https://arxiv.org/abs/2608.17607"}, "sourceUpdatedAt": "2026-08-18T10:16:33Z", "topics": ["Long Context", "Reasoning"], "venueAttempts": [], "whyItMatters": "Addresses the gap where final answer accuracy is insufficient to establish evidence grounding in pathology AI. Provides a protocol to assess whether models truly use supplied tissue evidence, offering a more rigorous evaluation for clinical deployment."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
