{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2606.12864", "kind": "paper", "url": "https://arxiv.org/abs/2606.12864"}], "authors": ["Tingqiang Xu", "Hangrui Zhou", "Tianle Cai", "Alex Gu", "Kaifeng Lyu"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.12864", "source_url": "https://arxiv.org/abs/2606.12864"}, "categories": ["Code", "Code & Software", "Code generation", "Software & AI Compute"], "description": {"en": "UOJ-Bench is a benchmark for evaluating LLMs in code generation, hacking, and repair, built from real-world submissions on the Universal Online Judge and evaluated through UOJ's judging infrastructure."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.12864", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.12864", "title": "UOJ-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Tsinghua University", "ByteDance (China)", "Massachusetts Institute of Technology"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2606.12864", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2606.12864", "source_url": "https://openalex.org/W7164608001"}, "key": "claire-radar:2606.12864", "modality": null, "name": "UOJ-Bench", "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_uoj-bench_c2013a7a", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "93e887a0cd40ec54cffc5fc9b0bd747835e03cf1fcd6426a2d32b4c968fc8de6", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2606.12864"}, "publisher": null, "released": "2026-06-11", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.12864", "source_url": "https://arxiv.org/abs/2606.12864"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-12864", "source": "claire_radar", "source_benchmark_id": "2606.12864", "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": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2606.12864", "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": ["Code generation"], "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": "2f39985c58bc9171a708ac301f999dd18b2e1ba063f6730c1b45475f188de916", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "UOJ-Bench is a benchmark for evaluating LLMs in code generation, hacking, and repair, built from real-world submissions on the Universal Online Judge and evaluated through UOJ's judging infrastructure.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact named benchmark artifact released in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "In this work, we introduce UOJ-Bench, a benchmark designed to evaluate not only the problem-solving ability of LLMs, but also their ability to identify errors in human-written code -- a crucial educational activity traditionally supported by running test cases over online judge systems."}, "familyId": "bmf_e38a83758c16", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_uoj-bench_c2013a7a", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": ["Software & Cloud"], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2606.12864", "paper": "https://arxiv.org/abs/2606.12864", "pdf": "https://arxiv.org/pdf/2606.12864", "project": null, "report": "https://arxiv.org/abs/2606.12864"}, "metrics": [], "motivation": "Despite strong performance in competitive programming, the role of Large Language Models (LLMs) in supporting human learning in the same setting remains largely unexplored.", "name": "UOJ-Bench", "oneLine": "UOJ-Bench is a benchmark for evaluating LLMs in code generation, hacking, and repair, built from real-world submissions on the Universal Online Judge and evaluated through UOJ's judging infrastructure.", "paperTitle": "Beyond Problem Solving: UOJ-Bench for Evaluating Code Generation, Hacking, and Repair in Competitive Programming", "primaryDomain": "Software & AI Compute", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2606.12864", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "ranking": {}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-06-11", "source": {"authors": ["Tingqiang Xu", "Hangrui Zhou", "Tianle Cai", "Alex Gu", "Kaifeng Lyu"], "categories": ["cs.SE", "cs.AI"], "id": "2606.12864", "title": "Beyond Problem Solving: UOJ-Bench for Evaluating Code Generation, Hacking, and Repair in Competitive Programming", "type": "arxiv", "url": "https://arxiv.org/abs/2606.12864"}, "sourceUpdatedAt": "2026-06-11T03:48:18Z", "topics": ["Code"], "venueAttempts": [], "whyItMatters": "UOJ-Bench extends beyond problem-solving to include identifying errors in human code, a critical educational activity. It provides a realistic setting for assessing LLM support in learning."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
