{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2607.27231", "kind": "paper", "url": "https://arxiv.org/abs/2607.27231"}, {"id": "gh:flagos-ai/kernelgenbench", "kind": "repo", "url": "https://github.com/flagos-ai/KernelGenBench"}], "authors": ["Peiyu Zang", "Jian Tao", "Jialing Zhang", "Yichen Yuan", "Wentao Zhang", "Guang Liu", "Yonghua Lin"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.27231", "source_url": "https://arxiv.org/abs/2607.27231"}, "categories": ["Agents", "General AI", "Language & Knowledge"], "description": {"en": "KernelGenBench evaluates LLM- and agent-generated Triton kernels across 210 operators from three sources (ATen, vLLM, cuBLAS) and six hardware platforms, with automatic accuracy verification and two evaluation tracks (LLM Pass@K and iterative agent generation)."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.27231", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.27231", "title": "KernelGenBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Beijing Normal University", "Beijing Jiaotong University", "Institute of Automation", "King University", "Peking University"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2607.27231", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2607.27231", "source_url": "https://openalex.org/W7171926671"}, "key": "claire-radar:2607.27231", "modality": null, "name": "KernelGenBench", "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_kernelgenbench_691527e5", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "9c039f3f45285e1efc2aadede1044560c7c2cf11175e94ab0e07b9742d63d60b", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2607.27231"}, "publisher": null, "released": "2026-07-22", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.27231", "source_url": "https://arxiv.org/abs/2607.27231"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-27231", "source": "claire_radar", "source_benchmark_id": "2607.27231", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/flagos-ai/KernelGenBench", "githubScope": "benchmark_repo", "githubStars": 20, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2607.27231", "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": "05d9b8c03fdb9a325e3dd3710de2b9132f4d6a92b07747954a68759928335d5c", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "KernelGenBench evaluates LLM- and agent-generated Triton kernels across 210 operators from three sources (ATen, vLLM, cuBLAS) and six hardware platforms, with automatic accuracy verification and two evaluation tracks (LLM Pass@K and iterative agent generation).", "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 present KernelGenBench, a unified benchmark for systematically evaluating LLM- and agent-generated Triton kernels across diverse operator sources and heterogeneous hardware platforms."}, "familyId": "bmf_12af6c0291ed", "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/flagos-ai/KernelGenBench", "url": "https://github.com/flagos-ai/KernelGenBench"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/flagos-ai/KernelGenBench", "url": "https://huggingface.co/papers/2607.27231"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/flagos-ai/KernelGenBench", "url": "https://arxiv.org/abs/2607.27231"}], "scope": "benchmark_repo", "url": "https://github.com/flagos-ai/KernelGenBench", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_kernelgenbench_691527e5", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/flagos-ai/KernelGenBench", "data": null, "hfPaper": "https://huggingface.co/papers/2607.27231", "paper": "https://arxiv.org/abs/2607.27231", "pdf": "https://arxiv.org/pdf/2607.27231", "project": null, "report": "https://arxiv.org/abs/2607.27231"}, "metrics": [], "motivation": "Large language models (LLMs) have significantly increased the demand for efficient accelerator kernels, but kernel development remains a highly specialized and labor-intensive task.", "name": "KernelGenBench", "oneLine": "KernelGenBench evaluates LLM- and agent-generated Triton kernels across 210 operators from three sources (ATen, vLLM, cuBLAS) and six hardware platforms, with automatic accuracy verification and two evaluation tracks (LLM Pass@K and iterative agent generation).", "paperTitle": "KernelGenBench: A Multi-Source and Multi-Chip Benchmark for LLM-based Kernel Generation", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.27231", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "FlagOS AI", "organizationType": "company-research-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/flagos-ai/KernelGenBench"}], "ranking": {"90d": {"confidence": "Low", "coverage": 0.55, "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": 0.4063890167320393, "value": 20}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 105, "score": 45}, "method": "attention-ranking-v13", "rank": 105, "score": 45}}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-07-22", "source": {"authors": ["Peiyu Zang", "Jian Tao", "Jialing Zhang", "Yichen Yuan", "Wentao Zhang", "Guang Liu", "and Yonghua Lin"], "categories": ["cs.AI", "cs.LG"], "id": "2607.27231", "title": "KernelGenBench: A Multi-Source and Multi-Chip Benchmark for LLM-based Kernel Generation", "type": "arxiv", "url": "https://arxiv.org/abs/2607.27231"}, "sourceUpdatedAt": "2026-07-22T12:40:11Z", "topics": ["Agents"], "venueAttempts": [], "whyItMatters": "Kernel generation is a specialized task lacking standardized evaluation; this benchmark provides a multi-source, multi-chip protocol to compare methods across diverse operators and hardware, enabling cost and portability assessment for autonomous kernel development."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
