{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2608.12004", "kind": "paper", "url": "https://arxiv.org/abs/2608.12004"}], "authors": ["Jinjun Huang", "Zhongzhen Wen", "Tongtong Xu", "Meng Yan", "Xin Xia", "Zhongxin Liu"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2608.12004", "source_url": "https://arxiv.org/abs/2608.12004"}, "categories": ["Code & Software", "Cybersecurity", "cs.SE"], "description": {"en": "RealisticTritonBench evaluates LLM-generated Triton kernels using tasks derived from real-world pull requests in popular AI frameworks, with end-to-end integration tests."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.12004", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.12004", "title": "RealisticTritonBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2608.12004", "modality": null, "name": "RealisticTritonBench", "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_realistictritonbench_fb357880", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "0c0cbd46350ff0b7f0fd2fd81f61527734e888a98cee8f02f4a342adaf91f1dc", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2608.12004"}, "publisher": null, "released": "2026-08-12", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.12004", "source_url": "https://arxiv.org/abs/2608.12004"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-12004", "source": "claire_radar", "source_benchmark_id": "2608.12004", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["Cybersecurity"], "area": "Code & Software", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2608.12004", "observedAt": "2026-10-04T03:02:42.754940Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "state": "unavailable"}}, "status": "unavailable"}, "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": "4245259960ec166ccdb4f1915bdc6d8dc983a5fcc26fa5b5e394d7e5c36877f0", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "RealisticTritonBench evaluates LLM-generated Triton kernels using tasks derived from real-world pull requests in popular AI frameworks, with end-to-end integration tests.", "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"], "snippet": "To address these limitations, we introduce RealisticTritonBench, the first benchmark to derive Triton kernel generation tasks from real-world pull requests in popular AI frameworks, enabling realistic, production-like evaluation."}, "familyId": "bmf_b1807ea10af3", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_realistictritonbench_fb357880", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": ["Cybersecurity"], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2608.12004", "paper": "https://arxiv.org/abs/2608.12004", "pdf": "https://arxiv.org/pdf/2608.12004", "project": null, "report": "https://arxiv.org/abs/2608.12004"}, "metrics": [], "motivation": "In modern AI frameworks, GPU kernels are key to overall system performance.", "name": "RealisticTritonBench", "oneLine": "RealisticTritonBench evaluates LLM-generated Triton kernels using tasks derived from real-world pull requests in popular AI frameworks, with end-to-end integration tests.", "paperTitle": "RealisticTritonBench: A Benchmark for Triton-Kernel Generation in Real-World AI Frameworks", "primaryDomain": "Cybersecurity", "publication": {"evidence": "Accepted by ASE 2026", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2608.12004", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "ASE 2026", "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "arXiv", "organizationType": "community", "role": "benchmark-publisher", "sourceUrl": "https://arxiv.org/abs/2608.12004"}], "ranking": {"90d": {"confidence": "Low", "coverage": 0.0, "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": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-08-12", "source": {"authors": ["Jinjun Huang", "Zhongzhen Wen", "Tongtong Xu", "Meng Yan", "Xin Xia", "Zhongxin Liu"], "categories": ["cs.SE", "cs.AI"], "id": "2608.12004", "title": "RealisticTritonBench: A Benchmark for Triton-Kernel Generation in Real-World AI Frameworks", "type": "arxiv", "url": "https://arxiv.org/abs/2608.12004"}, "sourceUpdatedAt": "2026-08-12T12:38:17Z", "topics": ["cs.SE"], "venueAttempts": [{"decisionRaw": "Accepted by ASE 2026", "evidence": [{"level": "author-claim", "observedAt": "2026-10-14T01:46:58.058199Z", "rawValue": "Accepted by ASE 2026", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2608.12004"}], "reviewStatus": "accepted", "venueName": "ASE 2026"}], "whyItMatters": "Existing benchmarks focus on isolated kernel translation and may have flawed evaluation scripts, while this benchmark provides realistic tasks and robust end-to-end evaluation, offering more practical insight into LLM performance for production kernel development."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
