{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2605.25388", "kind": "paper", "url": "https://arxiv.org/abs/2605.25388"}, {"id": "gh:qianjinydx/virobench", "kind": "repo", "url": "https://github.com/QIANJINYDX/ViroBench"}], "authors": ["Dongxin Ye", "Fang Hu", "Han Hu", "Shu Hu", "Yang Tan", "Wanli Ouyang", "Stan Z. Li", "Jie Cui", "Nanqing Dong"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2605.25388", "source_url": "https://arxiv.org/abs/2605.25388"}, "categories": ["Biology & Drug Discovery", "Language & Knowledge", "cs.LG"], "description": {"en": "ViroBench evaluates nucleotide foundation models on viral genomics tasks across two axes: biological understanding (taxonomy classification, host prediction) and latent biosecurity risk (genome modeling, CDS completion). It includes 18 tasks, 58,314 viral samples, multiple split strategies (genus-disjoint, temporal), and 11 metrics, with a unified evaluation protocol and leaderboard."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2605.25388", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2605.25388", "title": "ViroBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2605.25388", "modality": null, "name": "ViroBench", "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_virobench_ca7b2c7e", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "58e5abe56c82e81b868dd9e3dd372d732b1eb46f11fa1e3bdb9be1273d60e4e4", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2605.25388"}, "publisher": null, "released": "2026-05-25", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2605.25388", "source_url": "https://arxiv.org/abs/2605.25388"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2605-25388", "source": "claire_radar", "source_benchmark_id": "2605.25388", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["Biology & Drug Discovery"], "area": "Language & Knowledge", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/QIANJINYDX/ViroBench", "githubScope": "benchmark_repo", "githubStars": 15, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 0, "hfPaperUrl": "https://huggingface.co/papers/2605.25388", "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": "2026-09-30", "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": "922d8e86d3c93b75dd8274b046858b8510232f8630b545b8a53029625bb93f02", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "ViroBench evaluates nucleotide foundation models on viral genomics tasks across two axes: biological understanding (taxonomy classification, host prediction) and latent biosecurity risk (genome modeling, CDS completion). It includes 18 tasks, 58,314 viral samples, multiple split strategies (genus-disjoint, temporal), and 11 metrics, with a unified evaluation protocol and leaderboard.", "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": "To address this, we introduce ViroBench, the first comprehensive and large-scale benchmark specifically designed for NFMs in viral settings."}, "familyId": "bmf_3d16a3edd82a", "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/QIANJINYDX/ViroBench", "url": "https://github.com/QIANJINYDX/ViroBench"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/QIANJINYDX/ViroBench", "url": "https://huggingface.co/papers/2605.25388"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/QIANJINYDX/ViroBench", "url": "https://arxiv.org/abs/2605.25388"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/QIANJINYDX/ViroBench.", "url": "https://arxiv.org/abs/2605.25388"}], "scope": "benchmark_repo", "url": "https://github.com/QIANJINYDX/ViroBench", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_virobench_ca7b2c7e", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": ["Pharma & Biotech"], "links": {"code": "https://github.com/QIANJINYDX/ViroBench", "data": null, "hfPaper": "https://huggingface.co/papers/2605.25388", "paper": "https://arxiv.org/abs/2605.25388", "pdf": "https://arxiv.org/pdf/2605.25388", "project": null, "report": "https://arxiv.org/abs/2605.25388"}, "metrics": [], "motivation": "Nucleotide sequences constitute the fundamental genetic basis of biological systems, rendering viral genomic analysis critical for biomedical advancement.", "name": "ViroBench", "oneLine": "ViroBench evaluates nucleotide foundation models on viral genomics tasks across two axes: biological understanding (taxonomy classification, host prediction) and latent biosecurity risk (genome modeling, CDS completion). It includes 18 tasks, 58,314 viral samples, multiple split strategies (genus-disjoint, temporal), and 11 metrics, with a unified evaluation protocol and leaderboard.", "paperTitle": "ViroBench: Benchmarking Nucleotide Foundation Models on Viral Genomics Tasks", "primaryDomain": "Biology & Drug Discovery", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2605.25388", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "QIANJINYDX / ViroBench team", "organizationType": "community", "role": "benchmark-publisher", "sourceUrl": "https://github.com/QIANJINYDX/ViroBench"}], "ranking": {}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-05-25", "source": {"authors": ["Dongxin Ye", "Fang Hu", "Han Hu", "Shu Hu", "Yang Tan", "Wanli Ouyang", "Stan Z. Li", "Jie Cui", "Nanqing Dong"], "categories": ["cs.LG", "q-bio.QM"], "id": "2605.25388", "title": "ViroBench: Benchmarking Nucleotide Foundation Models on Viral Genomics Tasks", "type": "arxiv", "url": "https://arxiv.org/abs/2605.25388"}, "sourceUpdatedAt": "2026-05-25T03:31:46Z", "topics": ["cs.LG"], "venueAttempts": [], "whyItMatters": "Viral genomics lacks a unified evaluation standard for foundation models. ViroBench provides a diagnostic benchmark that measures both task performance and biosecurity risk, enabling model comparison and guiding development of safer, more robust viral nucleotide models."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
