[ { "URL": "https://arxiv.org/abs/2404.06474v3", "archive": "arXiv", "archive_location": "2404.06474", "author": [ { "literal": "Jiayi Pan" }, { "literal": "Yichi Zhang" }, { "literal": "Nicholas Tomlin" }, { "literal": "Yifei Zhou" }, { "literal": "Sergey Levine" }, { "literal": "Alane Suhr" } ], "citation-key": "arxiv_2404_06474", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "abc75ee55c7ac5992656dcec2106e286c670a58130a59b0eef7f5fd74d7496c5", "submitted_at": "2024-04-09T17:25:47Z", "updated_at": "2024-10-07T14:19:37Z", "version": 3, "versioned_url": "https://arxiv.org/abs/2404.06474v3" }, "canonical_source_url": "https://arxiv.org/abs/2404.06474", "catalog_id": "arxiv:2404.06474", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2024-04-09", "identifiers": { "arxiv": "2404.06474" }, "record_updated_at": "2024-10-07" }, "genre": "preprint", "id": "arxiv:2404.06474", "issued": { "date-parts": [ [ 2024, 4, 9 ] ] }, "note": "First published: 2024-04-09; Recorded arXiv version: v3; Recorded version updated: 2024-10-07T14:19:37Z.", "title": "Autonomous Evaluation and Refinement of Digital Agents", "type": "article", "version": "3" }, { "URL": "https://arxiv.org/abs/2411.03817v3", "archive": "arXiv", "archive_location": "2411.03817", "author": [ { "literal": "Zhirui Deng" }, { "literal": "Zhicheng Dou" }, { "literal": "Yutao Zhu" }, { "literal": "Ji-Rong Wen" }, { "literal": "Ruibin Xiong" }, { "literal": "Mang Wang" }, { "literal": "Weipeng Chen" } ], "citation-key": "arxiv_2411_03817", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "dd631a31f72dabe3a3bebf322031f9d686082073a8e4aa71eb65745498327f4e", "submitted_at": "2024-11-06T10:35:11Z", "updated_at": "2024-12-09T09:20:11Z", "version": 3, "versioned_url": "https://arxiv.org/abs/2411.03817v3" }, "canonical_source_url": "https://arxiv.org/abs/2411.03817", "catalog_id": "arxiv:2411.03817", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "adjacent", "first_published_at": "2024-11-06", "identifiers": { "arxiv": "2411.03817" }, "record_updated_at": "2024-12-09" }, "genre": "preprint", "id": "arxiv:2411.03817", "issued": { "date-parts": [ [ 2024, 11, 6 ] ] }, "note": "First published: 2024-11-06; Recorded arXiv version: v3; Recorded version updated: 2024-12-09T09:20:11Z.", "title": "From Novice to Expert: LLM Agent Policy Optimization via Step-wise Reinforcement Learning", "type": "article", "version": "3" }, { "URL": "https://arxiv.org/abs/2504.16073v1", "archive": "arXiv", "archive_location": "2504.16073", "author": [ { "literal": "Zhiyuan Hu" }, { "literal": "Shiyun Xiong" }, { "literal": "Yifan Zhang" }, { "literal": "See-Kiong Ng" }, { "literal": "Anh Tuan Luu" }, { "literal": "Bo An" }, { "literal": "Shuicheng Yan" }, { "literal": "Bryan Hooi" } ], "citation-key": "arxiv_2504_16073", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "1353bb4651f449763ac58803b4858edf6e1d5b0ad3ab3e4c333883a1ebbaebf4", "submitted_at": "2025-04-22T17:52:42Z", "updated_at": "2025-04-22T17:52:42Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2504.16073v1" }, "canonical_source_url": "https://arxiv.org/abs/2504.16073", "catalog_id": "arxiv:2504.16073", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-04-22", "identifiers": { "arxiv": "2504.16073" }, "record_updated_at": "2025-04-22" }, "genre": "preprint", "id": "arxiv:2504.16073", "issued": { "date-parts": [ [ 2025, 4, 22 ] ] }, "note": "First published: 2025-04-22; Recorded arXiv version: v1; Recorded version updated: 2025-04-22T17:52:42Z.", "title": "Guiding VLM Agents with Process Rewards at Inference Time for GUI Navigation", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2507.14293v1", "archive": "arXiv", "archive_location": "2507.14293", "author": [ { "literal": "Boyuan Zheng" }, { "literal": "Zeyi Liao" }, { "literal": "Scott Salisbury" }, { "literal": "Zeyuan Liu" }, { "literal": "Michael Lin" }, { "literal": "Qinyuan Zheng" }, { "literal": "Zifan Wang" }, { "literal": "Xiang Deng" }, { "literal": "Dawn Song" }, { "literal": "Huan Sun" }, { "literal": "Yu Su" } ], "citation-key": "arxiv_2507_14293", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "a8f082fd40174e01f40e97d64f7aa3d90e5f76e831df8f93afdc21aa2e35bfd5", "submitted_at": "2025-07-18T18:06:27Z", "updated_at": "2025-07-18T18:06:27Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2507.14293v1" }, "canonical_source_url": "https://arxiv.org/abs/2507.14293", "catalog_id": "arxiv:2507.14293", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-07-18", "identifiers": { "arxiv": "2507.14293" }, "record_updated_at": "2025-07-18" }, "genre": "preprint", "id": "arxiv:2507.14293", "issued": { "date-parts": [ [ 2025, 7, 18 ] ] }, "note": "First published: 2025-07-18; Recorded arXiv version: v1; Recorded version updated: 2025-07-18T18:06:27Z.", "title": "WebGuard: Building a Generalizable Guardrail for Web Agents", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2507.15846v3", "archive": "arXiv", "archive_location": "2507.15846", "author": [ { "literal": "Fei Tang" }, { "literal": "Zhangxuan Gu" }, { "literal": "Zhengxi Lu" }, { "literal": "Xuyang Liu" }, { "literal": "Shuheng Shen" }, { "literal": "Changhua Meng" }, { "literal": "Wen Wang" }, { "literal": "Wenqi Zhang" }, { "literal": "Yongliang Shen" }, { "literal": "Weiming Lu" }, { "literal": "Jun Xiao" }, { "literal": "Yueting Zhuang" } ], "citation-key": "arxiv_2507_15846", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "d2a2a49373259cb5903fad8c2fb48a7380732642bedd81a644e735af4f028c63", "submitted_at": "2025-07-21T17:53:42Z", "updated_at": "2025-07-28T16:54:13Z", "version": 3, "versioned_url": "https://arxiv.org/abs/2507.15846v3" }, "canonical_source_url": "https://arxiv.org/abs/2507.15846", "catalog_id": "arxiv:2507.15846", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-07-21", "identifiers": { "arxiv": "2507.15846" }, "record_updated_at": "2025-07-28" }, "genre": "preprint", "id": "arxiv:2507.15846", "issued": { "date-parts": [ [ 2025, 7, 21 ] ] }, "note": "First published: 2025-07-21; Recorded arXiv version: v3; Recorded version updated: 2025-07-28T16:54:13Z.", "title": "GUI-G$^2$: Gaussian Reward Modeling for GUI Grounding", "type": "article", "version": "3" }, { "URL": "https://arxiv.org/abs/2508.11360v2", "archive": "arXiv", "archive_location": "2508.11360", "author": [ { "literal": "Songqin Nong" }, { "literal": "Xiaoxuan Tang" }, { "literal": "Jingxuan Xu" }, { "literal": "Sheng Zhou" }, { "literal": "Jianfeng Chen" }, { "literal": "Tao Jiang" }, { "literal": "Wenhao Xu" } ], "citation-key": "arxiv_2508_11360", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "ea849d8e7640d7a40d582431233418cf1cbd29585ca6c78638919b8832265445", "submitted_at": "2025-08-15T09:55:02Z", "updated_at": "2026-03-13T11:10:50Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2508.11360v2" }, "canonical_source_url": "https://arxiv.org/abs/2508.11360", "catalog_id": "arxiv:2508.11360", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-08-15", "identifiers": { "arxiv": "2508.11360" }, "record_updated_at": "2026-03-13" }, "genre": "preprint", "id": "arxiv:2508.11360", "issued": { "date-parts": [ [ 2025, 8, 15 ] ] }, "note": "First published: 2025-08-15; Recorded arXiv version: v2; Recorded version updated: 2026-03-13T11:10:50Z.", "title": "CRAFT-GUI: Curriculum-Reinforced Agent For GUI Tasks", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2509.14172v2", "archive": "arXiv", "archive_location": "2509.14172", "author": [ { "literal": "Ziyuan Chen" }, { "literal": "Zhenghui Zhao" }, { "literal": "Zhangye Han" }, { "literal": "Miancan Liu" }, { "literal": "Xianhang Ye" }, { "literal": "Yiqing Li" }, { "literal": "Hongbo Min" }, { "literal": "Jinkui Ren" }, { "literal": "Xiantao Zhang" }, { "literal": "Guitao Cao" } ], "citation-key": "arxiv_2509_14172", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "cc1e4f78f1d3ffdb1b67f7c0bb645561a76db19dfa5f442fa9aa0a0d4d1ae0de", "submitted_at": "2025-09-17T16:58:44Z", "updated_at": "2025-09-19T02:13:09Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2509.14172v2" }, "canonical_source_url": "https://arxiv.org/abs/2509.14172", "catalog_id": "arxiv:2509.14172", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-09-17", "identifiers": { "arxiv": "2509.14172" }, "record_updated_at": "2025-09-19" }, "genre": "preprint", "id": "arxiv:2509.14172", "issued": { "date-parts": [ [ 2025, 9, 17 ] ] }, "note": "First published: 2025-09-17; Recorded arXiv version: v2; Recorded version updated: 2025-09-19T02:13:09Z.", "title": "TGPO: Tree-Guided Preference Optimization for Robust Web Agent Reinforcement Learning", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2509.17917v1", "archive": "arXiv", "archive_location": "2509.17917", "author": [ { "literal": "Junyu Lu" }, { "literal": "Songxin Zhang" }, { "literal": "Zejian Xie" }, { "literal": "Zhuoyang Song" }, { "literal": "Jiaxing Zhang" } ], "citation-key": "arxiv_2509_17917", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "7b85c9bae829bad343dd6f57c866ea100859fa0ad67f5c1a3ac9b3e53d617486", "submitted_at": "2025-09-22T15:40:31Z", "updated_at": "2025-09-22T15:40:31Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2509.17917v1" }, "canonical_source_url": "https://arxiv.org/abs/2509.17917", "catalog_id": "arxiv:2509.17917", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-09-22", "identifiers": { "arxiv": "2509.17917" }, "record_updated_at": "2025-09-22" }, "genre": "preprint", "id": "arxiv:2509.17917", "issued": { "date-parts": [ [ 2025, 9, 22 ] ] }, "note": "First published: 2025-09-22; Recorded arXiv version: v1; Recorded version updated: 2025-09-22T15:40:31Z.", "title": "Orcust: Stepwise-Feedback Reinforcement Learning for GUI Agent", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2509.21823v2", "archive": "arXiv", "archive_location": "2509.21823", "author": [ { "literal": "Gaole Dai" }, { "literal": "Shiqi Jiang" }, { "literal": "Ting Cao" }, { "literal": "Yuqing Yang" }, { "literal": "Yuanchun Li" }, { "literal": "Rui Tan" }, { "literal": "Mo Li" }, { "literal": "Lili Qiu" } ], "citation-key": "arxiv_2509_21823", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "c96f310fb0f28888c59dce71d08697b603e7a7562fdf99b51e3504b93e9e81b3", "submitted_at": "2025-09-26T03:29:36Z", "updated_at": "2026-04-15T09:19:46Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2509.21823v2" }, "canonical_source_url": "https://arxiv.org/abs/2509.21823", "catalog_id": "arxiv:2509.21823", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-09-26", "identifiers": { "arxiv": "2509.21823" }, "record_updated_at": "2026-04-15" }, "genre": "preprint", "id": "arxiv:2509.21823", "issued": { "date-parts": [ [ 2025, 9, 26 ] ] }, "note": "First published: 2025-09-26; Recorded arXiv version: v2; Recorded version updated: 2026-04-15T09:19:46Z.", "title": "ProRe: A Proactive Reward System for GUI Agents via Reasoner-Actor Collaboration", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2509.23263v3", "archive": "arXiv", "archive_location": "2509.23263", "author": [ { "literal": "Tao Xiong" }, { "literal": "Xavier Hu" }, { "literal": "Yurun Chen" }, { "literal": "Yuhang Liu" }, { "literal": "Changqiao Wu" }, { "literal": "Pengzhi Gao" }, { "literal": "Wei Liu" }, { "literal": "Jian Luan" }, { "literal": "Shengyu Zhang" } ], "citation-key": "arxiv_2509_23263", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "5b5375e77f087367383bb5de96cec46d8de6550d9a44bff6965b9201b6718a05", "submitted_at": "2025-09-27T11:42:36Z", "updated_at": "2026-08-30T18:12:14Z", "version": 3, "versioned_url": "https://arxiv.org/abs/2509.23263v3" }, "canonical_source_url": "https://arxiv.org/abs/2509.23263", "catalog_id": "arxiv:2509.23263", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-09-27", "identifiers": { "arxiv": "2509.23263" }, "record_updated_at": "2026-08-30" }, "genre": "preprint", "id": "arxiv:2509.23263", "issued": { "date-parts": [ [ 2025, 9, 27 ] ] }, "note": "First published: 2025-09-27; Recorded arXiv version: v3; Recorded version updated: 2026-08-30T18:12:14Z.", "title": "GUI-PRA: Process Reward Agent for GUI Tasks", "type": "article", "version": "3" }, { "URL": "https://arxiv.org/abs/2510.02250v2", "archive": "arXiv", "archive_location": "2510.02250", "author": [ { "literal": "Gonzalo Gonzalez-Pumariega" }, { "literal": "Vincent Tu" }, { "literal": "Chih-Lun Lee" }, { "literal": "Jiachen Yang" }, { "literal": "Ang Li" }, { "literal": "Xin Eric Wang" } ], "citation-key": "arxiv_2510_02250", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "29c00ba4713da26f912fec8046d498b4fd10cf71a3109f66e92fa23f7cebcf44", "submitted_at": "2025-10-02T17:37:08Z", "updated_at": "2026-02-03T23:48:45Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2510.02250v2" }, "canonical_source_url": "https://arxiv.org/abs/2510.02250", "catalog_id": "arxiv:2510.02250", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-10-02", "identifiers": { "arxiv": "2510.02250" }, "record_updated_at": "2026-02-03" }, "genre": "preprint", "id": "arxiv:2510.02250", "issued": { "date-parts": [ [ 2025, 10, 2 ] ] }, "note": "First published: 2025-10-02; Recorded arXiv version: v2; Recorded version updated: 2026-02-03T23:48:45Z.", "title": "Scaling Agents for Computer Use", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2512.10962v3", "archive": "arXiv", "archive_location": "2512.10962", "author": [ { "literal": "Yifei He" }, { "literal": "Pranit Chawla" }, { "literal": "Yaser Souri" }, { "literal": "Subhojit Som" }, { "literal": "Xia Song" } ], "citation-key": "arxiv_2512_10962", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "a7ccc7923f4e6f2ea296b4a6bfc28ba4dc7a52ed256008f153f13fccac3f018f", "submitted_at": "2025-11-22T23:12:56Z", "updated_at": "2026-02-04T19:36:03Z", "version": 3, "versioned_url": "https://arxiv.org/abs/2512.10962v3" }, "canonical_source_url": "https://arxiv.org/abs/2512.10962", "catalog_id": "arxiv:2512.10962", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-11-22", "identifiers": { "arxiv": "2512.10962" }, "record_updated_at": "2026-02-04" }, "genre": "preprint", "id": "arxiv:2512.10962", "issued": { "date-parts": [ [ 2025, 11, 22 ] ] }, "note": "First published: 2025-11-22; Recorded arXiv version: v3; Recorded version updated: 2026-02-04T19:36:03Z.", "title": "WebSTAR: Scalable Data Synthesis for Computer Use Agents with Step-Level Filtering", "type": "article", "version": "3" }, { "URL": "https://arxiv.org/abs/2512.15431v2", "archive": "arXiv", "archive_location": "2512.15431", "author": [ { "literal": "Haolong Yan" }, { "literal": "Jia Wang" }, { "literal": "Xin Huang" }, { "literal": "Yeqing Shen" }, { "literal": "Ziyang Meng" }, { "literal": "Zhimin Fan" }, { "literal": "Kaijun Tan" }, { "literal": "Jin Gao" }, { "literal": "Lieyu Shi" }, { "literal": "Mi Yang" }, { "literal": "Shiliang Yang" }, { "literal": "Zhirui Wang" }, { "literal": "Brian Li" }, { "literal": "Kang An" }, { "literal": "Chenyang Li" }, { "literal": "Lei Lei" }, { "literal": "Mengmeng Duan" }, { "literal": "Danxun Liang" }, { "literal": "Guodong Liu" }, { "literal": "Hang Cheng" }, { "literal": "Hao Wu" }, { "literal": "Jie Dong" }, { "literal": "Junhao Huang" }, { "literal": "Mei Chen" }, { "literal": "Renjie Yu" }, { "literal": "Shunshan Li" }, { "literal": "Xu Zhou" }, { "literal": "Yiting Dai" }, { "literal": "Yineng Deng" }, { "literal": "Yingdan Liang" }, { "literal": "Zelin Chen" }, { "literal": "Wen Sun" }, { "literal": "Chengxu Yan" }, { "literal": "Chunqin Xu" }, { "literal": "Dong Li" }, { "literal": "Fengqiong Xiao" }, { "literal": "Guanghao Fan" }, { "literal": "Guopeng Li" }, { "literal": "Guozhen Peng" }, { "literal": "Hongbing Li" }, { "literal": "Hang Li" }, { "literal": "Hongming Chen" }, { "literal": "Jingjing Xie" }, { "literal": "Jianyong Li" }, { "literal": "Jingyang Zhang" }, { "literal": "Jiaju Ren" }, { "literal": "Jiayu Yuan" }, { "literal": "Jianpeng Yin" }, { "literal": "Kai Cao" }, { "literal": "Liang Zhao" }, { "literal": "Liguo Tan" }, { "literal": "Liying Shi" }, { "literal": "Mengqiang Ren" }, { "literal": "Min Xu" }, { "literal": "Manjiao Liu" }, { "literal": "Mao Luo" }, { "literal": "Mingxin Wan" }, { "literal": "Na Wang" }, { "literal": "Nan Wu" }, { "literal": "Ning Wang" }, { "literal": "Peiyao Ma" }, { "literal": "Qingzhou Zhang" }, { "literal": "Qiao Wang" }, { "literal": "Qinlin Zeng" }, { "literal": "Qiong Gao" }, { "literal": "Qiongyao Li" }, { "literal": "Shangwu Zhong" }, { "literal": "Shuli Gao" }, { "literal": "Shaofan Liu" }, { "literal": "Shisi Gao" }, { "literal": "Shuang Luo" }, { "literal": "Xingbin Liu" }, { "literal": "Xiaojia Liu" }, { "literal": "Xiaojie Hou" }, { "literal": "Xin Liu" }, { "literal": "Xuanti Feng" }, { "literal": "Xuedan Cai" }, { "literal": "Xuan Wen" }, { "literal": "Xianwei Zhu" }, { "literal": "Xin Liang" }, { "literal": "Xin Liu" }, { "literal": "Xin Zhou" }, { "literal": "Yifan Sui" }, { "literal": "Yingxiu Zhao" }, { "literal": "Yukang Shi" }, { "literal": "Yunfang Xu" }, { "literal": "Yuqing Zeng" }, { "literal": "Yixun Zhang" }, { "literal": "Zejia Weng" }, { "literal": "Zhonghao Yan" }, { "literal": "Zhiguo Huang" }, { "literal": "Zhuoyu Wang" }, { "literal": "Zihan Yan" }, { "literal": "Zheng Ge" }, { "literal": "Jing Li" }, { "literal": "Yibo Zhu" }, { "literal": "Binxing Jiao" }, { "literal": "Xiangyu Zhang" }, { "literal": "Daxin Jiang" } ], "citation-key": "arxiv_2512_15431", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "2fc824bdeabca3451fd895447a0321c507da1ac7bc7cc24184c7768992549daf", "submitted_at": "2025-12-17T13:26:30Z", "updated_at": "2025-12-19T17:36:21Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2512.15431v2" }, "canonical_source_url": "https://arxiv.org/abs/2512.15431", "catalog_id": "arxiv:2512.15431", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-12-17", "identifiers": { "arxiv": "2512.15431" }, "record_updated_at": "2025-12-19" }, "genre": "preprint", "id": "arxiv:2512.15431", "issued": { "date-parts": [ [ 2025, 12, 17 ] ] }, "note": "First published: 2025-12-17; Recorded arXiv version: v2; Recorded version updated: 2025-12-19T17:36:21Z.", "title": "Step-GUI Technical Report", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2512.16295v1", "archive": "arXiv", "archive_location": "2512.16295", "author": [ { "literal": "Zhenyu Wu" }, { "literal": "Jingjing Xie" }, { "literal": "Zehao Li" }, { "literal": "Bowen Yang" }, { "literal": "Qiushi Sun" }, { "literal": "Zhaoyang Liu" }, { "literal": "Zhoumianze Liu" }, { "literal": "Yu Qiao" }, { "literal": "Xiangyu Yue" }, { "literal": "Zun Wang" }, { "literal": "Zichen Ding" } ], "citation-key": "arxiv_2512_16295", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "67a1bd40b117b34e5fb994b1c2c93cd285714097d1bc9730baeeaba6c31a7b8a", "submitted_at": "2025-12-18T08:29:50Z", "updated_at": "2025-12-18T08:29:50Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2512.16295v1" }, "canonical_source_url": "https://arxiv.org/abs/2512.16295", "catalog_id": "arxiv:2512.16295", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-12-18", "identifiers": { "arxiv": "2512.16295" }, "record_updated_at": "2025-12-18" }, "genre": "preprint", "id": "arxiv:2512.16295", "issued": { "date-parts": [ [ 2025, 12, 18 ] ] }, "note": "First published: 2025-12-18; Recorded arXiv version: v1; Recorded version updated: 2025-12-18T08:29:50Z.", "title": "OS-Oracle: A Comprehensive Framework for Cross-Platform GUI Critic Models", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2512.19396v3", "archive": "arXiv", "archive_location": "2512.19396", "author": [ { "literal": "Runze Li" }, { "literal": "Yuwen Zhai" }, { "literal": "Bo Xu" }, { "literal": "LiWu Xu" }, { "literal": "Nian Shi" }, { "literal": "Wei Zhang" }, { "literal": "Ran Lin" }, { "literal": "Liang Wang" } ], "citation-key": "arxiv_2512_19396", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "03cdf8514fa9f036b4614a880d842d49bb9cb0a821858b7215784fba209faa62", "submitted_at": "2025-12-22T13:42:18Z", "updated_at": "2026-04-10T06:20:00Z", "version": 3, "versioned_url": "https://arxiv.org/abs/2512.19396v3" }, "canonical_source_url": "https://arxiv.org/abs/2512.19396", "catalog_id": "arxiv:2512.19396", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2025-12-22", "identifiers": { "arxiv": "2512.19396" }, "record_updated_at": "2026-04-10" }, "genre": "preprint", "id": "arxiv:2512.19396", "issued": { "date-parts": [ [ 2025, 12, 22 ] ] }, "note": "First published: 2025-12-22; Recorded arXiv version: v3; Recorded version updated: 2026-04-10T06:20:00Z.", "title": "EchoTrail-GUI: Building Actionable Memory for GUI Agents via Critic-Guided Self-Exploration", "type": "article", "version": "3" }, { "URL": "https://arxiv.org/abs/2601.13060v1", "archive": "arXiv", "archive_location": "2601.13060", "author": [ { "literal": "Zecheng Li" }, { "literal": "Zhihui Cao" }, { "literal": "Wenke Huang" }, { "literal": "Yudong Zhang" }, { "literal": "Keying Qi" }, { "literal": "Rui Wang" }, { "literal": "Zeyu Zheng" }, { "literal": "Jian Zhao" }, { "literal": "Hao Zhu" }, { "literal": "Hengxin Wu" }, { "literal": "Yuran Wang" }, { "literal": "Guitao Fan" }, { "literal": "Guokun Wu" }, { "literal": "Yicong Liu" }, { "literal": "Zhilin Gao" }, { "literal": "Haikun Xu" }, { "literal": "He Yang" }, { "literal": "Minqi Xiang" }, { "literal": "Xingyu Liu" }, { "literal": "Zuojian Wang" } ], "citation-key": "arxiv_2601_13060", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "aa8cb4137e862bf8031a55bc8f81e1d6c903733f2cfac9cf758e5f2f92884481", "submitted_at": "2026-01-19T13:50:43Z", "updated_at": "2026-01-19T13:50:43Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2601.13060v1" }, "canonical_source_url": "https://arxiv.org/abs/2601.13060", "catalog_id": "arxiv:2601.13060", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-01-19", "identifiers": { "arxiv": "2601.13060" }, "record_updated_at": "2026-01-19" }, "genre": "preprint", "id": "arxiv:2601.13060", "issued": { "date-parts": [ [ 2026, 1, 19 ] ] }, "note": "First published: 2026-01-19; Recorded arXiv version: v1; Recorded version updated: 2026-01-19T13:50:43Z.", "title": "MagicGUI-RMS: A Multi-Agent Reward Model System for Self-Evolving GUI Agents via Automated Feedback Reflux", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2601.18197v2", "archive": "arXiv", "archive_location": "2601.18197", "author": [ { "literal": "Shaokang Wang" }, { "literal": "Pei Fu" }, { "literal": "Ruoceng Zhang" }, { "literal": "Shaojie Zhang" }, { "literal": "Xiuwen Xi" }, { "literal": "Jiahui Yang" }, { "literal": "Bin Qin" }, { "literal": "Ying Huang" }, { "literal": "Zhenbo Luo" }, { "literal": "Jian Luan" } ], "citation-key": "arxiv_2601_18197", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "ccb46b86b0c0b854ed2f1e17f9e12bf031c5ed1349f3d3c44d3276b2a3a195ee", "submitted_at": "2026-01-26T06:29:41Z", "updated_at": "2026-06-26T07:19:55Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2601.18197v2" }, "canonical_source_url": "https://arxiv.org/abs/2601.18197", "catalog_id": "arxiv:2601.18197", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-01-26", "identifiers": { "arxiv": "2601.18197" }, "record_updated_at": "2026-06-26" }, "genre": "preprint", "id": "arxiv:2601.18197", "issued": { "date-parts": [ [ 2026, 1, 26 ] ] }, "note": "First published: 2026-01-26; Recorded arXiv version: v2; Recorded version updated: 2026-06-26T07:19:55Z.", "title": "GAIA: A Data Flywheel System for Training GUI Test-Time Scaling Critic Models", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2601.21872v2", "archive": "arXiv", "archive_location": "2601.21872", "author": [ { "literal": "Yao Zhang" }, { "literal": "Shijie Tang" }, { "literal": "Zeyu Li" }, { "literal": "Zhen Han" }, { "literal": "Volker Tresp" } ], "citation-key": "arxiv_2601_21872", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "68ae67856eb871d9a8f259dfb180bde02d29ee97fd7abf0e60719fc2c7abd7ab", "submitted_at": "2026-01-29T15:39:50Z", "updated_at": "2026-04-09T17:48:11Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2601.21872v2" }, "canonical_source_url": "https://arxiv.org/abs/2601.21872", "catalog_id": "arxiv:2601.21872", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-01-29", "identifiers": { "arxiv": "2601.21872" }, "record_updated_at": "2026-04-09" }, "genre": "preprint", "id": "arxiv:2601.21872", "issued": { "date-parts": [ [ 2026, 1, 29 ] ] }, "note": "First published: 2026-01-29; Recorded arXiv version: v2; Recorded version updated: 2026-04-09T17:48:11Z.", "title": "WebArbiter: A Principle-Guided Reasoning Process Reward Model for Web Agents", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2602.00575v2", "archive": "arXiv", "archive_location": "2602.00575", "author": [ { "literal": "Chaoqun Cui" }, { "literal": "Jing Huang" }, { "literal": "Shijing Wang" }, { "literal": "Liming Zheng" }, { "literal": "Qingchao Kong" }, { "literal": "Zhixiong Zeng" } ], "citation-key": "arxiv_2602_00575", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "86225f2e52e1394e2a0a56b1e82327d756bff95739e01ce7cc5898af49944aba", "submitted_at": "2026-01-31T07:36:54Z", "updated_at": "2026-07-25T06:54:35Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2602.00575v2" }, "canonical_source_url": "https://arxiv.org/abs/2602.00575", "catalog_id": "arxiv:2602.00575", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-01-31", "identifiers": { "arxiv": "2602.00575" }, "record_updated_at": "2026-07-25" }, "genre": "preprint", "id": "arxiv:2602.00575", "issued": { "date-parts": [ [ 2026, 1, 31 ] ] }, "note": "First published: 2026-01-31; Recorded arXiv version: v2; Recorded version updated: 2026-07-25T06:54:35Z.", "title": "Agentic Reward Modeling: Verifying GUI Agent via Progressive Trajectory-Grounded Interaction", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2602.10356v2", "archive": "arXiv", "archive_location": "2602.10356", "author": [ { "literal": "Tianci Xue" }, { "literal": "Zeyi Liao" }, { "literal": "Tianneng Shi" }, { "literal": "Zilu Wang" }, { "literal": "Kai Zhang" }, { "literal": "Dawn Song" }, { "literal": "Yu Su" }, { "literal": "Huan Sun" } ], "citation-key": "arxiv_2602_10356", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "d3385966e7b483b33738b3c8e6c9a4d9cf17afa671e1f2e49de4f946cbc5f251", "submitted_at": "2026-02-10T23:06:02Z", "updated_at": "2026-05-11T00:11:46Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2602.10356v2" }, "canonical_source_url": "https://arxiv.org/abs/2602.10356", "catalog_id": "arxiv:2602.10356", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-02-10", "identifiers": { "arxiv": "2602.10356" }, "record_updated_at": "2026-05-11" }, "genre": "preprint", "id": "arxiv:2602.10356", "issued": { "date-parts": [ [ 2026, 2, 10 ] ] }, "note": "First published: 2026-02-10; Recorded arXiv version: v2; Recorded version updated: 2026-05-11T00:11:46Z.", "title": "Autonomous Continual Learning for Environment Adaptation of Computer-Use Agents", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2602.11524v1", "archive": "arXiv", "archive_location": "2602.11524", "author": [ { "literal": "Congmin Zheng" }, { "literal": "Xiaoyun Mo" }, { "literal": "Xinbei Ma" }, { "literal": "Qiqiang Lin" }, { "literal": "Yin Zhao" }, { "literal": "Jiachen Zhu" }, { "literal": "Xingyu Lou" }, { "literal": "Jun Wang" }, { "literal": "Zhaoxiang Wang" }, { "literal": "Weiwen Liu" }, { "literal": "Zhuosheng Zhang" }, { "literal": "Yong Yu" }, { "literal": "Weinan Zhang" } ], "citation-key": "arxiv_2602_11524", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "ffec5a47e31b541cfdf0fb3c143b55cbb28119463731d9a6f3add0fe116caea9", "submitted_at": "2026-02-12T03:31:40Z", "updated_at": "2026-02-12T03:31:40Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2602.11524v1" }, "canonical_source_url": "https://arxiv.org/abs/2602.11524", "catalog_id": "arxiv:2602.11524", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-02-12", "identifiers": { "arxiv": "2602.11524" }, "record_updated_at": "2026-02-12" }, "genre": "preprint", "id": "arxiv:2602.11524", "issued": { "date-parts": [ [ 2026, 2, 12 ] ] }, "note": "First published: 2026-02-12; Recorded arXiv version: v1; Recorded version updated: 2026-02-12T03:31:40Z.", "title": "Adaptive Milestone Reward for GUI Agents", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2602.13653v1", "archive": "arXiv", "archive_location": "2602.13653", "author": [ { "literal": "Yibo Wang" }, { "literal": "Guangda Huzhang" }, { "literal": "Yuwei Hu" }, { "literal": "Yu Xia" }, { "literal": "Shiyin Lu" }, { "literal": "Qing-Guo Chen" }, { "literal": "Zhao Xu" }, { "literal": "Weihua Luo" }, { "literal": "Kaifu Zhang" }, { "literal": "Lijun Zhang" } ], "citation-key": "arxiv_2602_13653", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "573d859133ec9986af92e1943d0ec913d914acf4e54a030a4b103181ba505ae0", "submitted_at": "2026-02-14T07:44:47Z", "updated_at": "2026-02-14T07:44:47Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2602.13653v1" }, "canonical_source_url": "https://arxiv.org/abs/2602.13653", "catalog_id": "arxiv:2602.13653", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-02-14", "identifiers": { "arxiv": "2602.13653" }, "record_updated_at": "2026-02-14" }, "genre": "preprint", "id": "arxiv:2602.13653", "issued": { "date-parts": [ [ 2026, 2, 14 ] ] }, "note": "First published: 2026-02-14; Recorded arXiv version: v1; Recorded version updated: 2026-02-14T07:44:47Z.", "title": "Building Autonomous GUI Navigation via Agentic-Q Estimation and Step-Wise Policy Optimization", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2603.10178v1", "archive": "arXiv", "archive_location": "2603.10178", "author": [ { "literal": "Linxin Song" }, { "literal": "Jieyu Zhang" }, { "literal": "Huanxin Sheng" }, { "literal": "Taiwei Shi" }, { "literal": "Gupta Rahul" }, { "literal": "Yang Liu" }, { "literal": "Ranjay Krishna" }, { "literal": "Jian Kang" }, { "literal": "Jieyu Zhao" } ], "citation-key": "arxiv_2603_10178", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "22b3001540ac8cc75117cdfb4f381462fe7df3fe7700a86e14b9f522803316ac", "submitted_at": "2026-03-10T19:17:22Z", "updated_at": "2026-03-10T19:17:22Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2603.10178v1" }, "canonical_source_url": "https://arxiv.org/abs/2603.10178", "catalog_id": "arxiv:2603.10178", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-03-10", "identifiers": { "arxiv": "2603.10178" }, "record_updated_at": "2026-03-10" }, "genre": "preprint", "id": "arxiv:2603.10178", "issued": { "date-parts": [ [ 2026, 3, 10 ] ] }, "note": "First published: 2026-03-10; Recorded arXiv version: v1; Recorded version updated: 2026-03-10T19:17:22Z.", "title": "Video-Based Reward Modeling for Computer-Use Agents", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2603.19191v1", "archive": "arXiv", "archive_location": "2603.19191", "author": [ { "literal": "Zehao Li" }, { "literal": "Zhenyu Wu" }, { "literal": "Yibo Zhao" }, { "literal": "Bowen Yang" }, { "literal": "Jingjing Xie" }, { "literal": "Zhaoyang Liu" }, { "literal": "Zhoumianze Liu" }, { "literal": "Kaiming Jin" }, { "literal": "Jianze Liang" }, { "literal": "Zonglin Li" }, { "literal": "Feng Wu" }, { "literal": "Bowen Zhou" }, { "literal": "Zun Wang" }, { "literal": "Zichen Ding" } ], "citation-key": "arxiv_2603_19191", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "ff31371580af6a0c43a6c26f076d5060ee06f974d3a6c1ade67348dc0af4859d", "submitted_at": "2026-03-19T17:47:47Z", "updated_at": "2026-03-19T17:47:47Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2603.19191v1" }, "canonical_source_url": "https://arxiv.org/abs/2603.19191", "catalog_id": "arxiv:2603.19191", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-03-19", "identifiers": { "arxiv": "2603.19191" }, "record_updated_at": "2026-03-19" }, "genre": "preprint", "id": "arxiv:2603.19191", "issued": { "date-parts": [ [ 2026, 3, 19 ] ] }, "note": "First published: 2026-03-19; Recorded arXiv version: v1; Recorded version updated: 2026-03-19T17:47:47Z.", "title": "OS-Themis: A Scalable Critic Framework for Generalist GUI Rewards", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2603.19685v1", "archive": "arXiv", "archive_location": "2603.19685", "author": [ { "literal": "Taiyi Wang" }, { "literal": "Sian Gooding" }, { "literal": "Florian Hartmann" }, { "literal": "Oriana Riva" }, { "literal": "Edward Grefenstette" } ], "citation-key": "arxiv_2603_19685", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "5538933a0e3d11df6ff4d767849979703d83412599fbbb0f8b432c4d9850633b", "submitted_at": "2026-03-20T06:32:16Z", "updated_at": "2026-03-20T06:32:16Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2603.19685v1" }, "canonical_source_url": "https://arxiv.org/abs/2603.19685", "catalog_id": "arxiv:2603.19685", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-03-20", "identifiers": { "arxiv": "2603.19685" }, "record_updated_at": "2026-03-20" }, "genre": "preprint", "id": "arxiv:2603.19685", "issued": { "date-parts": [ [ 2026, 3, 20 ] ] }, "note": "First published: 2026-03-20; Recorded arXiv version: v1; Recorded version updated: 2026-03-20T06:32:16Z.", "title": "A Subgoal-driven Framework for Improving Long-Horizon LLM Agents", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2604.05157v3", "archive": "arXiv", "archive_location": "2604.05157", "author": [ { "literal": "Rongqian Chen" }, { "literal": "Yu Li" }, { "literal": "Zeyu Fang" }, { "literal": "Sizhe Tang" }, { "literal": "Weidong Cao" }, { "literal": "Tian Lan" } ], "citation-key": "arxiv_2604_05157", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "238aac264fb21022f7c546ae3aa3f4ea104c381f5d318f6d962f22fb00095216", "submitted_at": "2026-04-06T20:39:30Z", "updated_at": "2026-05-28T08:13:55Z", "version": 3, "versioned_url": "https://arxiv.org/abs/2604.05157v3" }, "canonical_source_url": "https://arxiv.org/abs/2604.05157", "catalog_id": "arxiv:2604.05157", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-04-06", "identifiers": { "arxiv": "2604.05157" }, "record_updated_at": "2026-05-28" }, "genre": "preprint", "id": "arxiv:2604.05157", "issued": { "date-parts": [ [ 2026, 4, 6 ] ] }, "note": "First published: 2026-04-06; Recorded arXiv version: v3; Recorded version updated: 2026-05-28T08:13:55Z.", "title": "IntentScore: Intent-Conditioned Action Evaluation for Computer-Use Agents", "type": "article", "version": "3" }, { "URL": "https://arxiv.org/abs/2604.06240v1", "archive": "arXiv", "archive_location": "2604.06240", "author": [ { "literal": "Corby Rosset" }, { "literal": "Pratyusha Sharma" }, { "literal": "Andrew Zhao" }, { "literal": "Miguel Gonzalez-Fernandez" }, { "literal": "Ahmed Awadallah" } ], "citation-key": "arxiv_2604_06240", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "b452684b305c4f20ded686289b448fbe4f08bb3b34f515a43e2ce5d6dc4ef506", "submitted_at": "2026-04-05T06:26:05Z", "updated_at": "2026-04-05T06:26:05Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2604.06240v1" }, "canonical_source_url": "https://arxiv.org/abs/2604.06240", "catalog_id": "arxiv:2604.06240", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-04-05", "identifiers": { "arxiv": "2604.06240" }, "record_updated_at": "2026-04-05" }, "genre": "preprint", "id": "arxiv:2604.06240", "issued": { "date-parts": [ [ 2026, 4, 5 ] ] }, "note": "First published: 2026-04-05; Recorded arXiv version: v1; Recorded version updated: 2026-04-05T06:26:05Z.", "title": "The Art of Building Verifiers for Computer Use Agents", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2604.07277v2", "archive": "arXiv", "archive_location": "2604.07277", "author": [ { "literal": "Guo Gan" }, { "literal": "Yuxuan Ding" }, { "literal": "Cong Chen" }, { "literal": "Yuwei Ren" }, { "literal": "Yin Huang" }, { "literal": "Hong Zhou" } ], "citation-key": "arxiv_2604_07277", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "f4d0dbeac950760f4f676ab7c651a881fd275f4adb69d46a94a11c4f10a03980", "submitted_at": "2026-04-08T16:40:26Z", "updated_at": "2026-05-08T03:06:31Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2604.07277v2" }, "canonical_source_url": "https://arxiv.org/abs/2604.07277", "catalog_id": "arxiv:2604.07277", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-04-08", "identifiers": { "arxiv": "2604.07277" }, "record_updated_at": "2026-05-08" }, "genre": "preprint", "id": "arxiv:2604.07277", "issued": { "date-parts": [ [ 2026, 4, 8 ] ] }, "note": "First published: 2026-04-08; Recorded arXiv version: v2; Recorded version updated: 2026-05-08T03:06:31Z.", "title": "Android Coach: Improve Online Agentic Training Efficiency with Single State Multiple Actions", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2604.18240v1", "archive": "arXiv", "archive_location": "2604.18240", "author": [ { "literal": "Wentao Shi" }, { "literal": "Yu Wang" }, { "literal": "Yuyang Zhao" }, { "literal": "Yuxin Chen" }, { "literal": "Fuli Feng" }, { "literal": "Xueyuan Hao" }, { "literal": "Xi Su" }, { "literal": "Qi Gu" }, { "literal": "Hui Su" }, { "literal": "Xunliang Cai" }, { "literal": "Xiangnan He" } ], "citation-key": "arxiv_2604_18240", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "91a9cdb6c8cb124d2e6033f0decc85e0c73bd972984d973ff687bbdc2be21176", "submitted_at": "2026-04-20T13:23:38Z", "updated_at": "2026-04-20T13:23:38Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2604.18240v1" }, "canonical_source_url": "https://arxiv.org/abs/2604.18240", "catalog_id": "arxiv:2604.18240", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-04-20", "identifiers": { "arxiv": "2604.18240" }, "record_updated_at": "2026-04-20" }, "genre": "preprint", "id": "arxiv:2604.18240", "issued": { "date-parts": [ [ 2026, 4, 20 ] ] }, "note": "First published: 2026-04-20; Recorded arXiv version: v1; Recorded version updated: 2026-04-20T13:23:38Z.", "title": "AJ-Bench: Benchmarking Agent-as-a-Judge for Environment-Aware Evaluation", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2606.07027v2", "archive": "arXiv", "archive_location": "2606.07027", "author": [ { "literal": "Haojie Hao" }, { "literal": "Longkun Hao" }, { "literal": "Yihang Lou" }, { "literal": "Yan Bai" }, { "literal": "Zhenyang Li" }, { "literal": "Zhichao Yang" }, { "literal": "Dongshuo Huang" }, { "literal": "Hongyu Lin" }, { "literal": "Lanqing Hong" }, { "literal": "Jiakai Wang" }, { "literal": "Xianglong Liu" } ], "citation-key": "arxiv_2606_07027", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "2d530152b210a69cb458fecb036c10621c72daccfbcac7c8306c7e248c87cf14", "submitted_at": "2026-06-05T08:17:28Z", "updated_at": "2026-06-12T07:28:19Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2606.07027v2" }, "canonical_source_url": "https://arxiv.org/abs/2606.07027", "catalog_id": "arxiv:2606.07027", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-06-05", "identifiers": { "arxiv": "2606.07027" }, "record_updated_at": "2026-06-12" }, "genre": "preprint", "id": "arxiv:2606.07027", "issued": { "date-parts": [ [ 2026, 6, 5 ] ] }, "note": "First published: 2026-06-05; Recorded arXiv version: v2; Recorded version updated: 2026-06-12T07:28:19Z.", "title": "StainFlow: Entity-Stain Tracking and Evidence Linking for Process Rewards in GUI Agents", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2606.11078v1", "archive": "arXiv", "archive_location": "2606.11078", "author": [ { "literal": "Jaewoo Lee" }, { "literal": "Zaid Khan" }, { "literal": "Archiki Prasad" }, { "literal": "Justin Chih-Yao Chen" }, { "literal": "Supriyo Chakraborty" }, { "literal": "Kartik Balasubramaniam" }, { "literal": "Sambit Sahu" }, { "literal": "Elias Stengel-Eskin" }, { "literal": "Hyunji Lee" }, { "literal": "Mohit Bansal" } ], "citation-key": "arxiv_2606_11078", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "a8ed399c8b489a80c5313a00232d2ecca0fbc8228b1f56745c02d4fedb6f6908", "submitted_at": "2026-06-09T16:39:10Z", "updated_at": "2026-06-09T16:39:10Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2606.11078v1" }, "canonical_source_url": "https://arxiv.org/abs/2606.11078", "catalog_id": "arxiv:2606.11078", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-06-09", "identifiers": { "arxiv": "2606.11078" }, "record_updated_at": "2026-06-09" }, "genre": "preprint", "id": "arxiv:2606.11078", "issued": { "date-parts": [ [ 2026, 6, 9 ] ] }, "note": "First published: 2026-06-09; Recorded arXiv version: v1; Recorded version updated: 2026-06-09T16:39:10Z.", "title": "A History-Aware Visually Grounded Critic for Computer Use Agents", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2606.24515v1", "archive": "arXiv", "archive_location": "2606.24515", "author": [ { "literal": "Marta Sumyk" }, { "literal": "Oleksandr Kosovan" } ], "citation-key": "arxiv_2606_24515", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "0aeb8c28b570133f575c2f336b6dab0dcbecdfaec78549115407b21722816e8a", "submitted_at": "2026-06-23T12:46:29Z", "updated_at": "2026-06-23T12:46:29Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2606.24515v1" }, "canonical_source_url": "https://arxiv.org/abs/2606.24515", "catalog_id": "arxiv:2606.24515", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-06-23", "identifiers": { "arxiv": "2606.24515" }, "record_updated_at": "2026-06-23" }, "genre": "preprint", "id": "arxiv:2606.24515", "issued": { "date-parts": [ [ 2026, 6, 23 ] ] }, "note": "First published: 2026-06-23; Recorded arXiv version: v1; Recorded version updated: 2026-06-23T12:46:29Z.", "title": "Reinforcement Learning for Computer-Use Agents with Autonomous Evaluation", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2606.24525v1", "archive": "arXiv", "archive_location": "2606.24525", "author": [ { "literal": "Jiachen Qian" } ], "citation-key": "arxiv_2606_24525", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "a3d0341f41a8ecba57fb7ed9e4b266bb4a176639c742be2b4fc98555754f5744", "submitted_at": "2026-06-23T12:57:02Z", "updated_at": "2026-06-23T12:57:02Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2606.24525v1" }, "canonical_source_url": "https://arxiv.org/abs/2606.24525", "catalog_id": "arxiv:2606.24525", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-06-23", "identifiers": { "arxiv": "2606.24525" }, "record_updated_at": "2026-06-23" }, "genre": "preprint", "id": "arxiv:2606.24525", "issued": { "date-parts": [ [ 2026, 6, 23 ] ] }, "note": "First published: 2026-06-23; Recorded arXiv version: v1; Recorded version updated: 2026-06-23T12:57:02Z.", "title": "VisCritic: Visual State Comparison as Process Reward for GUI Agents", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2607.23263v2", "archive": "arXiv", "archive_location": "2607.23263", "author": [ { "literal": "Yang Wan" }, { "literal": "Zhenhao Zhang" }, { "literal": "Jierui Wang" }, { "literal": "Xiangxin Zhou" }, { "literal": "Linchao Zhu" } ], "citation-key": "arxiv_2607_23263", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "2de35aa3c7f4457d43d485eb136d8b380ea61870cc7253e807880ae731c51316", "submitted_at": "2026-07-25T16:00:45Z", "updated_at": "2026-09-26T12:15:44Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2607.23263v2" }, "canonical_source_url": "https://arxiv.org/abs/2607.23263", "catalog_id": "arxiv:2607.23263", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-07-25", "identifiers": { "arxiv": "2607.23263" }, "record_updated_at": "2026-09-26" }, "genre": "preprint", "id": "arxiv:2607.23263", "issued": { "date-parts": [ [ 2026, 7, 25 ] ] }, "note": "First published: 2026-07-25; Recorded arXiv version: v2; Recorded version updated: 2026-09-26T12:15:44Z.", "title": "SeekJudge: A Practical Reward Framework for Reinforcement Learning in Computer-Use Agents", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2607.25904v2", "archive": "arXiv", "archive_location": "2607.25904", "author": [ { "literal": "Chenrui Shi" }, { "literal": "Yuwei Wu" }, { "literal": "Yang Liu" }, { "literal": "Ruining Feng" }, { "literal": "Zirui Shang" }, { "literal": "Zhi Gao" }, { "literal": "Lifeng Fan" }, { "literal": "Che Sun" } ], "citation-key": "arxiv_2607_25904", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "d1688a528d35a8213a8027385b1b1327536567f6ec307aa6562eea543ee71d7f", "submitted_at": "2026-07-28T16:01:38Z", "updated_at": "2026-07-29T05:13:57Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2607.25904v2" }, "canonical_source_url": "https://arxiv.org/abs/2607.25904", "catalog_id": "arxiv:2607.25904", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-07-28", "identifiers": { "arxiv": "2607.25904" }, "record_updated_at": "2026-07-29" }, "genre": "preprint", "id": "arxiv:2607.25904", "issued": { "date-parts": [ [ 2026, 7, 28 ] ] }, "note": "First published: 2026-07-28; Recorded arXiv version: v2; Recorded version updated: 2026-07-29T05:13:57Z.", "title": "Interactive Reward Agent: GUI Task Evaluation via Environment-State Verification", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2607.28609v2", "archive": "arXiv", "archive_location": "2607.28609", "author": [ { "literal": "Qiushi Sun" }, { "literal": "Kanzhi Cheng" }, { "literal": "Yian Wang" }, { "literal": "Bowen Yang" }, { "literal": "Hang Yan" }, { "literal": "Liheng Chen" }, { "literal": "Fangzhi Xu" }, { "literal": "Zichen Ding" }, { "literal": "Nuo Chen" }, { "literal": "Jialin Cao" }, { "literal": "Xingdong Gong" }, { "literal": "Zehao Li" }, { "literal": "Kaiming Jin" }, { "literal": "Xinfeng Yuan" }, { "literal": "Zhoumianze Liu" }, { "literal": "Jingyang Gong" }, { "literal": "Zhangyue Yin" }, { "literal": "Jiahui Gao" }, { "literal": "Zhiyong Wu" }, { "literal": "Tianbao Xie" }, { "literal": "Jianbing Zhang" }, { "literal": "Ben Kao" }, { "literal": "Lingpeng Kong" } ], "citation-key": "arxiv_2607_28609", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "e3e82df7abffbd0c8f7e969ed8edb77423263a2b1d826a3e72529b0a51903d58", "submitted_at": "2026-07-30T17:57:41Z", "updated_at": "2026-08-06T17:55:59Z", "version": 2, "versioned_url": "https://arxiv.org/abs/2607.28609v2" }, "canonical_source_url": "https://arxiv.org/abs/2607.28609", "catalog_id": "arxiv:2607.28609", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-07-30", "identifiers": { "arxiv": "2607.28609" }, "record_updated_at": "2026-08-06" }, "genre": "preprint", "id": "arxiv:2607.28609", "issued": { "date-parts": [ [ 2026, 7, 30 ] ] }, "note": "First published: 2026-07-30; Recorded arXiv version: v2; Recorded version updated: 2026-08-06T17:55:59Z.", "title": "OSReward: Instituting Standardized Evaluation for Cross-Platform Computer-Use Reward Models", "type": "article", "version": "2" }, { "URL": "https://arxiv.org/abs/2608.05587v1", "archive": "arXiv", "archive_location": "2608.05587", "author": [ { "literal": "Linqiang Guo" }, { "literal": "Wei Liu" }, { "literal": "Li Gu" }, { "literal": "Yang Wang" }, { "literal": "Tse-Hsun" }, { "literal": "Chen" } ], "citation-key": "arxiv_2608_05587", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "dc7934927c82a43e6ae6a246ade1d50c4551cc49d263ca67cfb32906cbec9ea4", "submitted_at": "2026-08-06T04:17:27Z", "updated_at": "2026-08-06T04:17:27Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2608.05587v1" }, "canonical_source_url": "https://arxiv.org/abs/2608.05587", "catalog_id": "arxiv:2608.05587", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-08-06", "identifiers": { "arxiv": "2608.05587" }, "record_updated_at": "2026-08-06" }, "genre": "preprint", "id": "arxiv:2608.05587", "issued": { "date-parts": [ [ 2026, 8, 6 ] ] }, "note": "First published: 2026-08-06; Recorded arXiv version: v1; Recorded version updated: 2026-08-06T04:17:27Z.", "title": "StepReflect: Structured UI Transition Reflection for Mobile GUI Agents", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2608.11434v1", "archive": "arXiv", "archive_location": "2608.11434", "author": [ { "literal": "Ziqiang Wang" }, { "literal": "Li Gu" }, { "literal": "Zhixiang Chi" }, { "literal": "Zhi Liu" }, { "literal": "Seyed Mehdi Ayyoubzadeh" }, { "literal": "Yuanhao Yu" }, { "literal": "Yang Wang" } ], "citation-key": "arxiv_2608_11434", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "a2606ae96ca3ea99b5c7278485601f9bde3b64eefdd3bfecc2bbfc678c39b4fc", "submitted_at": "2026-08-11T21:00:46Z", "updated_at": "2026-08-11T21:00:46Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2608.11434v1" }, "canonical_source_url": "https://arxiv.org/abs/2608.11434", "catalog_id": "arxiv:2608.11434", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-08-11", "identifiers": { "arxiv": "2608.11434" }, "record_updated_at": "2026-08-11" }, "genre": "preprint", "id": "arxiv:2608.11434", "issued": { "date-parts": [ [ 2026, 8, 11 ] ] }, "note": "First published: 2026-08-11; Recorded arXiv version: v1; Recorded version updated: 2026-08-11T21:00:46Z.", "title": "Benchmarking LLM Judges for Mobile Agent Evaluation", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2608.22847v1", "archive": "arXiv", "archive_location": "2608.22847", "author": [ { "literal": "Long Zhang" }, { "literal": "Yuhan Chen" }, { "literal": "Chaoran Zhang" }, { "literal": "Wanxia Cao" }, { "literal": "Kun Huang" }, { "literal": "Pengzhi Gao" }, { "literal": "Wei Liu" }, { "literal": "Jian Luan" }, { "literal": "Chenliang Li" }, { "literal": "Lixin Zou" } ], "citation-key": "arxiv_2608_22847", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "4c1198ce65d5a2fcffc51b37bc4553b35b9e0d809757dc4181747c57c7ae443d", "submitted_at": "2026-08-24T06:27:47Z", "updated_at": "2026-08-24T06:27:47Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2608.22847v1" }, "canonical_source_url": "https://arxiv.org/abs/2608.22847", "catalog_id": "arxiv:2608.22847", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-08-24", "identifiers": { "arxiv": "2608.22847" }, "record_updated_at": "2026-08-24" }, "genre": "preprint", "id": "arxiv:2608.22847", "issued": { "date-parts": [ [ 2026, 8, 24 ] ] }, "note": "First published: 2026-08-24; Recorded arXiv version: v1; Recorded version updated: 2026-08-24T06:27:47Z.", "title": "GSAR: Goal-State-Anchor Rewards for Mobile GUI Agents with Self-Evolving Data Synthesis", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2608.24174v1", "archive": "arXiv", "archive_location": "2608.24174", "author": [ { "literal": "Tao Xiong" }, { "literal": "Xavier Hu" }, { "literal": "Wenkai Wang" }, { "literal": "Qinzhuo Wu" }, { "literal": "Changqiao Wu" }, { "literal": "Pengzhi Gao" }, { "literal": "Wei Liu" }, { "literal": "Jian Luan" }, { "literal": "Shengyu Zhang" } ], "citation-key": "arxiv_2608_24174", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "76cd09f70bfec0d2a49cddae8adcf32ca37fb39bcade7c1d36d5f3dc3fa6bfa7", "submitted_at": "2026-08-25T07:40:12Z", "updated_at": "2026-08-25T07:40:12Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2608.24174v1" }, "canonical_source_url": "https://arxiv.org/abs/2608.24174", "catalog_id": "arxiv:2608.24174", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-08-25", "identifiers": { "arxiv": "2608.24174" }, "record_updated_at": "2026-08-25" }, "genre": "preprint", "id": "arxiv:2608.24174", "issued": { "date-parts": [ [ 2026, 8, 25 ] ] }, "note": "First published: 2026-08-25; Recorded arXiv version: v1; Recorded version updated: 2026-08-25T07:40:12Z.", "title": "Task-Adaptive Rubrics for GUI Reward Modeling", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2609.02885v1", "archive": "arXiv", "archive_location": "2609.02885", "author": [ { "literal": "Kelvin Li" }, { "literal": "Dhruv Pendharkar" }, { "literal": "Anish Pahilajani" }, { "literal": "Chuyi Shang" }, { "literal": "Leon Oks" }, { "literal": "Leonid Karlinsky" }, { "literal": "Rogerio Feris" }, { "literal": "Trevor Darrell" }, { "literal": "Roei Herzig" } ], "citation-key": "arxiv_2609_02885", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "8703474cb4f577b307731cf1a986f76f800470fde625b6ca4b3569d30e74f903", "submitted_at": "2026-09-02T17:59:40Z", "updated_at": "2026-09-02T17:59:40Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2609.02885v1" }, "canonical_source_url": "https://arxiv.org/abs/2609.02885", "catalog_id": "arxiv:2609.02885", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-09-02", "identifiers": { "arxiv": "2609.02885" }, "record_updated_at": "2026-09-02" }, "genre": "preprint", "id": "arxiv:2609.02885", "issued": { "date-parts": [ [ 2026, 9, 2 ] ] }, "note": "First published: 2026-09-02; Recorded arXiv version: v1; Recorded version updated: 2026-09-02T17:59:40Z.", "title": "Discriminative World Models for Web Agents", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2609.32351v1", "archive": "arXiv", "archive_location": "2609.32351", "author": [ { "literal": "Yangzhe Peng" }, { "literal": "Xiaoyang Wang" }, { "literal": "Yiyang Zhao" }, { "literal": "Lijun Wu" }, { "literal": "Kun He" } ], "citation-key": "arxiv_2609_32351", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "7cc6fd87d2299e2881eab9a2db96e4022663e2d94da5549dcde166c0c8b8be11", "submitted_at": "2026-09-26T08:14:50Z", "updated_at": "2026-09-26T08:14:50Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2609.32351v1" }, "canonical_source_url": "https://arxiv.org/abs/2609.32351", "catalog_id": "arxiv:2609.32351", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-09-26", "identifiers": { "arxiv": "2609.32351" }, "record_updated_at": "2026-09-26" }, "genre": "preprint", "id": "arxiv:2609.32351", "issued": { "date-parts": [ [ 2026, 9, 26 ] ] }, "note": "First published: 2026-09-26; Recorded arXiv version: v1; Recorded version updated: 2026-09-26T08:14:50Z.", "title": "From Trajectories to Grounded Preferences: Process Preference Synthesis via Interaction Element Graphs for Web PRMs", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2610.18067v1", "archive": "arXiv", "archive_location": "2610.18067", "author": [ { "literal": "Hang He" }, { "literal": "Linting Wang" }, { "literal": "Guanming Liu" }, { "literal": "Bo Mao" }, { "literal": "Jinxiu Liu" }, { "literal": "Yiran Suo" }, { "literal": "Maosen Zhou" }, { "literal": "Zhiheng Xi" }, { "literal": "Guanting Dong" }, { "literal": "Zhou Shao" }, { "literal": "Rui Zheng" }, { "literal": "Jie Zhou" }, { "literal": "Chengcheng Wan" }, { "literal": "Tao Gui" }, { "literal": "Qi Zhang" }, { "literal": "Xipeng Qiu" }, { "literal": "Xuanjing Huang" } ], "citation-key": "arxiv_2610_18067", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "6add363f25f87ea1814c7b79ebf6b51f7431b3745a7d2d910960b78e2d97d13f", "submitted_at": "2026-10-13T09:52:02Z", "updated_at": "2026-10-13T09:52:02Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2610.18067v1" }, "canonical_source_url": "https://arxiv.org/abs/2610.18067", "catalog_id": "arxiv:2610.18067", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-10-13", "identifiers": { "arxiv": "2610.18067" }, "record_updated_at": "2026-10-13" }, "genre": "preprint", "id": "arxiv:2610.18067", "issued": { "date-parts": [ [ 2026, 10, 13 ] ] }, "note": "First published: 2026-10-13; Recorded arXiv version: v1; Recorded version updated: 2026-10-13T09:52:02Z.", "title": "WorkRover: Scaling Long-Horizon Cowork Agent Training Via Recursive Environment Evolution", "type": "article", "version": "1" }, { "URL": "https://arxiv.org/abs/2610.26469v1", "archive": "arXiv", "archive_location": "2610.26469", "author": [ { "literal": "Zhizhou Wang" }, { "literal": "Yaxi Lu" }, { "literal": "Zhong Zhang" }, { "literal": "Zikun Hu" }, { "literal": "Xin Cong" }, { "literal": "Yankai Lin" }, { "literal": "Fangming Liu" } ], "citation-key": "arxiv_2610_26469", "custom": { "arxiv_snapshot": { "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1", "abstract_sha256": "08ef7c9cdda652f60d3f2d4ab89e7218ad5a2b85f5c76f0277c289db06fd2796", "submitted_at": "2026-10-19T14:17:33Z", "updated_at": "2026-10-19T14:17:33Z", "version": 1, "versioned_url": "https://arxiv.org/abs/2610.26469v1" }, "canonical_source_url": "https://arxiv.org/abs/2610.26469", "catalog_id": "arxiv:2610.26469", "catalog_kind": "paper", "catalog_review_status": "curated", "catalog_scope": "core", "first_published_at": "2026-10-19", "identifiers": { "arxiv": "2610.26469" }, "record_updated_at": "2026-10-19" }, "genre": "preprint", "id": "arxiv:2610.26469", "issued": { "date-parts": [ [ 2026, 10, 19 ] ] }, "note": "First published: 2026-10-19; Recorded arXiv version: v1; Recorded version updated: 2026-10-19T14:17:33Z.", "title": "VeriCUA: Learning Proactive Verification for Reliable GUI Agent Rewards", "type": "article", "version": "1" } ]