{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2609.08149", "kind": "paper", "url": "https://arxiv.org/abs/2609.08149"}], "authors": ["Pujun Zheng", "Zixin Shang", "Shufan Jiang", "Wenhui Tian", "Dongsheng Zhu", "Zerun Ma", "Dingbo Yuan", "Qi Zhang"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.08149", "source_url": "https://arxiv.org/abs/2609.08149"}, "categories": ["Code", "Code & Software", "Software & AI Compute"], "description": {"en": "A verified version of SWE-Bench Pro for evaluating software engineering agents on repository-level coding tasks."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.08149", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.08149", "title": "SWE-Bench Pro Verified"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Shanghai Artificial Intelligence Laboratory", "East China Normal University", "Fudan University"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2609.08149", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2609.08149", "source_url": "https://openalex.org/W7212059813"}, "key": "claire-radar:2609.08149", "modality": null, "name": "SWE-Bench Pro Verified", "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": "false", "origin_record_id": "bm_swe-bench_6e4ae275", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "aa6ea390ab4d435f25d224fdd75f3ffe0f706300b33d69187fac05ebd3995263", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-name-audit-deferred", "reviewed_at": "2026-09-09T06:27:57.312318Z", "source_url": "https://arxiv.org/abs/2609.08149"}, "publisher": null, "released": "2026-09-08", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.08149", "source_url": "https://arxiv.org/abs/2609.08149"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-08149", "source": "claire_radar", "source_benchmark_id": "2609.08149", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["Software & AI Compute"], "area": "Code & Software", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": "2026-09-10T00:00:00.000Z", "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 27, "hfPaperUrl": "https://huggingface.co/papers/2609.08149", "observedAt": "2026-10-04T15:31:48.030916Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}}, "status": "fresh"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "The topic of reliable agent evaluation could draw interest, but no release evidence was provided.", "score": 30}, "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-09T06:27:57.312318Z", "inputHash": "fa69b8b3fcdf3f037fff6ffe5f8dad171aef9d58537b213dd5227735187b5ca6", "model": "deepseek-v4-pro"}, "curation": {"decisionReason": "The paper describes a benchmark derivation but supplies no evidence of released task instances, scoring implementation, or public submission path.", "model": "deepseek-v4-pro", "policy": "source-evidence-v2", "reviewedAt": "2026-09-09T06:27:57.312318Z", "state": "ai-name-audit-deferred"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "A verified version of SWE-Bench Pro for evaluating software engineering agents on repository-level coding tasks.", "displayEligible": false, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["exact named benchmark artifact released in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We present \\textbf{SWE-Bench Pro Verified}, a verified version of SWE-Bench Pro that addresses both problems."}, "familyId": "bmf_bd016a717f69", "firstSeenAt": "2026-09-09", "heat": null, "id": "bm_swe-bench_6e4ae275", "indexedAt": "2026-09-09T06:18:29.794093Z", "industrySectors": ["Software & Cloud"], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2609.08149", "pdf": "https://arxiv.org/pdf/2609.08149", "project": null, "report": "https://arxiv.org/abs/2609.08149"}, "metrics": [], "motivation": "SWE-Bench Pro has emerged as a standard benchmark for evaluating software engineering agents on challenging repository-level tasks.", "name": "SWE-Bench Pro Verified", "oneLine": "A verified version of SWE-Bench Pro for evaluating software engineering agents on repository-level coding tasks.", "paperTitle": "SWE-Bench Pro Verified: A Reliable Benchmark for Software Engineering Agents", "primaryDomain": "Software & AI Compute", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2609.08149", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-09T06:18:29.794093Z"}, "publications": [], "ranking": {"30d": {"confidence": "Low", "coverage": 0.3, "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": 30, "historyDate": "2026-09-05", "rank": null, "score": null, "windowDays": 30}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": 0.620448537533797, "value": 27}}, "confidence": "Low", "coverage": 0.3, "rank": null, "score": 54}, "method": "attention-ranking-v13", "rank": null, "score": 54}, "90d": {"confidence": "Low", "coverage": 0.15, "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": 0.5887825746490047, "value": 27}}, "confidence": "Low", "coverage": 0.15, "rank": null, "score": 51}, "method": "attention-ranking-v13", "rank": null, "score": 51}}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-09-08", "source": {"authors": ["Pujun Zheng", "Zixin Shang", "Shufan Jiang", "Wenhui Tian", "Dongsheng Zhu", "Zerun Ma", "Dingbo Yuan and Qi Zhang"], "categories": ["cs.AI", "cs.SE"], "id": "2609.08149", "title": "SWE-Bench Pro Verified: A Reliable Benchmark for Software Engineering Agents", "type": "arxiv", "url": "https://arxiv.org/abs/2609.08149"}, "sourceUpdatedAt": "2026-09-08T02:29:58Z", "topics": ["Code"], "venueAttempts": [], "whyItMatters": "Addresses reliability issues such as reward hacking and flawed tasks to provide more trustworthy measurements of agent capability."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
