{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2608.04783", "kind": "paper", "url": "https://arxiv.org/abs/2608.04783"}, {"id": "gh:tencent-hunyuan/repoprobe", "kind": "repo", "url": "https://github.com/Tencent-Hunyuan/RepoProbe"}], "authors": ["Yuexi Yang", "Alyssa Wu", "Ji Luo", "Richeng Xuan", "Zhichao Hu", "Yuhong Liu", "Zhen Qin"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2608.04783", "source_url": "https://arxiv.org/abs/2608.04783"}, "categories": ["Code & Software", "Software & AI Compute", "cs.SE"], "description": {"en": "Evaluates repository-level code understanding via 500 open-ended questions derived from GitHub Discussions across 50 repositories, scored with checklist-based verification for factual correctness and clarity."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.04783", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.04783", "title": "RepoProbe"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2608.04783", "modality": null, "name": "RepoProbe", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "Medium", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "true", "origin_record_id": "bm_repoprobe_4f581859", "origin_source": "arxiv", "recognition_confidence": "0.8", "record_sha256": "ade086fd52742842e9fda05e4af51a71d4e79c49782cfaf80156105bf621c642", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-08T08:57:27.463024Z", "source_url": "https://arxiv.org/abs/2608.04783"}, "publisher": null, "released": "2026-08-05", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.04783", "source_url": "https://arxiv.org/abs/2608.04783"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-04783", "source": "claire_radar", "source_benchmark_id": "2608.04783", "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": "https://github.com/Tencent-Hunyuan/RepoProbe", "githubScope": "benchmark_repo", "githubStars": 13, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 0, "hfPaperUrl": "https://huggingface.co/papers/2608.04783", "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": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}}, "status": "fresh"}, "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "Accepted to a prestigious software engineering conference and provides a code and dataset release, likely attracting attention from LLM evaluation and software engineering researchers.", "score": 85}, "capabilities": [], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-08T08:57:27.463024Z", "inputHash": "07b66a4b6ffbfc31169cdeef8dba9b91a29b52dd29200afcdc1b310d3867b94c", "model": "deepseek-v4-pro"}, "curation": {"canonicalNameEvidence": "RepoProbe: Benchmarking Architecture-Aware Repository Comprehension with Checklists", "canonicalNameSource": "paper_title", "decisionReason": "Formal benchmark with a clear scoring protocol, released replication package, and explicit evaluation methodology suitable for model comparison.", "model": "deepseek-v4-pro", "reviewedAt": "2026-09-08T08:57:27.463024Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates repository-level code understanding via 500 open-ended questions derived from GitHub Discussions across 50 repositories, scored with checklist-based verification for factual correctness and clarity.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "score_submission", "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "public artifact URL"], "snippet": "This work introduces RepoProbe, a novel benchmark for evaluating repository-level code understanding through open-ended Q&A using GitHub Discussions, which focuses on open-ended architectural inquiries rather than defect reporting."}, "familyId": "bmf_c7b217d2f7a8", "firstSeenAt": "2026-08-19", "heat": null, "id": "bm_repoprobe_4f581859", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": ["Software & Cloud"], "links": {"code": "https://github.com/Tencent-Hunyuan/RepoProbe", "data": null, "paper": "https://arxiv.org/abs/2608.04783", "pdf": "https://arxiv.org/pdf/2608.04783", "project": null, "report": "https://arxiv.org/abs/2608.04783"}, "metrics": [], "motivation": "The integration of Large Language Models (LLMs) into software engineering has shifted the focus from function-level generation to repository-scale assistance.", "name": "RepoProbe", "oneLine": "Evaluates repository-level code understanding via 500 open-ended questions derived from GitHub Discussions across 50 repositories, scored with checklist-based verification for factual correctness and clarity.", "paperTitle": "RepoProbe: Benchmarking Architecture-Aware Repository Comprehension with Checklists", "primaryDomain": "Software & AI Compute", "publication": {"evidence": "Accepted to the 41st IEEE/ACM International Conference on Automated Software Engineering (ASE 2026). Replication package: https://github.com/Tencent-Hunyuan/RepoProbe", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2608.04783", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "41st IEEE/ACM International Conference on Automated Software Engineering (ASE 2026)", "verifiedAt": "2026-08-19T11:05:13.395391Z"}, "publications": [], "publishers": [{"name": "Tencent Hunyuan", "organizationType": "company-research-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/Tencent-Hunyuan/RepoProbe"}], "ranking": {"90d": {"confidence": "Medium", "coverage": 0.7, "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": 0.3522667134894853, "value": 13}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": 0.0, "value": 0}}, "confidence": "Medium", "coverage": 0.7, "rank": 194, "score": 34}, "method": "attention-ranking-v13", "rank": 194, "score": 34}}, "readiness": "Runnable", "recognitionConfidence": 0.8, "relation": "introduces", "releasedAt": "2026-08-05", "source": {"authors": ["Yuexi Yang", "Alyssa Wu", "Ji Luo", "Richeng Xuan", "Zhichao Hu", "Yuhong Liu", "Zhen Qin"], "categories": ["cs.SE", "cs.AI"], "id": "2608.04783", "title": "RepoProbe: Benchmarking Architecture-Aware Repository Comprehension with Checklists", "type": "arxiv", "url": "https://arxiv.org/abs/2608.04783"}, "sourceUpdatedAt": "2026-08-06T09:16:21Z", "topics": ["cs.SE"], "venueAttempts": [{"decisionRaw": "Accepted to the 41st IEEE/ACM International Conference on Automated Software Engineering (ASE 2026). Replication package: https://github.com/Tencent-Hunyuan/RepoProbe", "evidence": [{"level": "author-claim", "observedAt": "2026-08-19T11:05:13.395391Z", "rawValue": "Accepted to the 41st IEEE/ACM International Conference on Automated Software Engineering (ASE 2026). Replication package: https://github.com/Tencent-Hunyuan/RepoProbe", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2608.04783"}], "reviewStatus": "accepted", "venueName": "41st IEEE/ACM International Conference on Automated Software Engineering (ASE 2026)"}], "whyItMatters": "Addresses limitations of defect-centric benchmarks that allow shortcut pattern matching by focusing on architectural comprehension, providing objective item-level scoring that improves reliability over scalar LLM-as-a-Judge ratings."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
