{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2606.11042", "kind": "paper", "url": "https://arxiv.org/abs/2606.11042"}], "authors": ["Liya Zhu", "Jingzhe Ding", "Jian Zhang", "Jianbo Xue", "Shihao Liang", "Ge Zhang", "Yi Zhu", "Duju Zeng", "Xiang Gao", "Qingshui Gu", "Mailun Gao", "Huimin Che", "Yan Zhao", "Peiheng Zhou", "Haojun Wang", "Chaobo Xian", "Lili Le", "Chi Wu", "Yiwei Liu", "Shengda Long", "Jiale Yang", "Fangzhi Xu", "Sijin Wu", "Haodong Duan", "Chao He", "Zhaojian Li", "Minchao Wang", "Huan Zhou", "Jiani Hou", "Chuqian Yu", "Weiran Shi", "Hongwan Gao", "Jiamin Chen", "Guanhong Chen", "Tingqin Luo", "Kaiyuan Zhang", "Zhixin Yao", "Qing Hua", "Yuhao Jiang", "Jin Chen", "Pu Chen", "Zhenyu Hu", "Xingyu Li", "Zhengxuan Jiang", "Meng Cao", "Tianfeng Long", "Haozhe Wang", "Mingzhang Wang", "Yichen Zhang", "Yiming Dai", "Chenchen Zhang", "Jiaying Wang", "Xinying Liu", "Xingzu Liu", "Lingling Zhang", "Xinjie Chen", "Yujia Qin", "Wangchunshu Zhou", "Zhiyong Wu", "Yang Liu", "Jiaheng Liu", "Lei Zhang", "Shen Yan", "Wenhao Huang", "Zaiyuan Wang", "Xiaolong Chang"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.11042", "source_url": "https://arxiv.org/abs/2606.11042"}, "categories": ["General AI", "Language & Knowledge", "cs.AI"], "description": {"en": "Workflow-GYM evaluates AI agents on long-horizon GUI tasks in professional domains, using specialized software environments and economically valuable workflows. Tasks require end-to-end operation of graphical user interfaces, with success rates measured by task completion."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.11042", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.11042", "title": "Workflow-GYM"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2606.11042", "modality": null, "name": "Workflow-GYM", "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_workflow-gym_72fcdcd4", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "e9427b59acc82c7c1d2104cb0797178e7d713f51c0cb6c45f5704351786e0c3d", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2606.11042"}, "publisher": null, "released": "2026-06-09", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.11042", "source_url": "https://arxiv.org/abs/2606.11042"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-11042", "source": "claire_radar", "source_benchmark_id": "2606.11042", "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": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 23, "hfPaperUrl": "https://huggingface.co/papers/2606.11042", "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", "lastSuccessfulDate": "2026-10-01", "state": "stale"}}, "status": "stale"}, "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": "badacf3a03dd44823e3abf9dd0292873ffea1d6474fd00c6cb1085e6cffd3063", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Workflow-GYM evaluates AI agents on long-horizon GUI tasks in professional domains, using specialized software environments and economically valuable workflows. Tasks require end-to-end operation of graphical user interfaces, with success rates measured by task completion.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "To bridge this gap, we introduce Workflow-GYM, a benchmark for long-horizon GUI tasks centered on professional domains and specialized software environments."}, "familyId": "bmf_bc38808cdab5", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_workflow-gym_72fcdcd4", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2606.11042", "paper": "https://arxiv.org/abs/2606.11042", "pdf": "https://arxiv.org/pdf/2606.11042", "project": null, "report": "https://arxiv.org/abs/2606.11042"}, "metrics": [], "motivation": "Recent years have witnessed the rapid evolution of AI agents toward handling increasingly complex, real-world tasks.", "name": "Workflow-GYM", "oneLine": "Workflow-GYM evaluates AI agents on long-horizon GUI tasks in professional domains, using specialized software environments and economically valuable workflows. Tasks require end-to-end operation of graphical user interfaces, with success rates measured by task completion.", "paperTitle": "Workflow-GYM: Towards Long-Horizon Evaluation of Computer-use Agentic tasks in Real-World Professional Fields", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2606.11042", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "Workflow-GYM Team", "organizationType": "academic-lab", "role": "benchmark-publisher", "sourceUrl": "https://arxiv.org/abs/2606.11042"}], "ranking": {}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-06-09", "source": {"authors": ["Liya Zhu", "Jingzhe Ding", "Jian Zhang", "Jianbo Xue", "Shihao Liang", "Ge Zhang", "Yi Zhu", "Duju Zeng", "Xiang Gao", "Qingshui Gu", "Mailun Gao", "Huimin Che", "Yan Zhao", "Peiheng Zhou", "Haojun Wang", "Chaobo Xian", "Lili Le", "Chi Wu", "Yiwei Liu", "Shengda Long", "Jiale Yang", "Fangzhi Xu", "Sijin Wu", "Haodong Duan", "Chao He", "Zhaojian Li", "Minchao Wang", "Huan Zhou", "Jiani Hou", "Chuqian Yu", "Weiran Shi", "Hongwan Gao", "Jiamin Chen", "Guanhong Chen", "Tingqin Luo", "Kaiyuan Zhang", "Zhixin Yao", "Qing Hua", "Yuhao Jiang", "Jin Chen", "Pu Chen", "Zhenyu Hu", "Xingyu Li", "Zhengxuan Jiang", "Meng Cao", "Tianfeng Long", "Haozhe Wang", "Mingzhang Wang", "Yichen Zhang", "Yiming Dai", "Chenchen Zhang", "Jiaying Wang", "Xinying Liu", "Xingzu Liu", "Lingling Zhang", "Xinjie Chen", "Yujia Qin", "Wangchunshu Zhou", "Zhiyong Wu", "Yang Liu", "Jiaheng Liu", "Lei Zhang", "Shen Yan", "Wenhao Huang", "Zaiyuan Wang", "Xiaolong Chang"], "categories": ["cs.AI"], "id": "2606.11042", "title": "Workflow-GYM: Towards Long-Horizon Evaluation of Computer-use Agentic tasks in Real-World Professional Fields", "type": "arxiv", "url": "https://arxiv.org/abs/2606.11042"}, "sourceUpdatedAt": "2026-07-17T12:22:12Z", "topics": ["cs.AI"], "venueAttempts": [], "whyItMatters": "Most GUI benchmarks cover simple, short-horizon tasks in general software, leaving a gap for professional, long-horizon workflows. Workflow-GYM provides a fixed protocol for measuring agent performance on such tasks, which is valuable as organizations consider deploying agents for complex professional work."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
