[
  {
    "URL": "https://arxiv.org/abs/1602.02261v2",
    "archive": "arXiv",
    "archive_location": "1602.02261",
    "author": [
      {
        "literal": "Rodrigo Nogueira"
      },
      {
        "literal": "Kyunghyun Cho"
      }
    ],
    "citation-key": "arxiv_1602_02261",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "67f8697c21dabc2c4781eba46bfb60671506b413471efcdb28864c11649b3a2c",
        "submitted_at": "2016-02-06T14:53:02Z",
        "updated_at": "2016-05-20T16:26:58Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/1602.02261v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/1602.02261",
      "catalog_id": "arxiv:1602.02261",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2016-02-06",
      "identifiers": {
        "arxiv": "1602.02261"
      },
      "record_updated_at": "2016-05-20"
    },
    "genre": "preprint",
    "id": "arxiv:1602.02261",
    "issued": {
      "date-parts": [
        [
          2016,
          2,
          6
        ]
      ]
    },
    "note": "First published: 2016-02-06; Recorded arXiv version: v2; Recorded version updated: 2016-05-20T16:26:58Z.",
    "title": "End-to-End Goal-Driven Web Navigation",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2305.14257v3",
    "archive": "arXiv",
    "archive_location": "2305.14257",
    "author": [
      {
        "literal": "Abishek Sridhar"
      },
      {
        "literal": "Robert Lo"
      },
      {
        "literal": "Frank F. Xu"
      },
      {
        "literal": "Hao Zhu"
      },
      {
        "literal": "Shuyan Zhou"
      }
    ],
    "citation-key": "arxiv_2305_14257",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "896d5f5e9248e2e49da254fb383f84e5b6ce4286b6d30a37ed7944fcfa1f4958",
        "submitted_at": "2023-05-23T17:10:39Z",
        "updated_at": "2023-10-30T00:55:53Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2305.14257v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2305.14257",
      "catalog_id": "arxiv:2305.14257",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2023-05-23",
      "identifiers": {
        "arxiv": "2305.14257"
      },
      "record_updated_at": "2023-10-30"
    },
    "genre": "preprint",
    "id": "arxiv:2305.14257",
    "issued": {
      "date-parts": [
        [
          2023,
          5,
          23
        ]
      ]
    },
    "note": "First published: 2023-05-23; Recorded arXiv version: v3; Recorded version updated: 2023-10-30T00:55:53Z.",
    "title": "Hierarchical Prompting Assists Large Language Model on Web Navigation",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2305.16653v1",
    "archive": "arXiv",
    "archive_location": "2305.16653",
    "author": [
      {
        "literal": "Haotian Sun"
      },
      {
        "literal": "Yuchen Zhuang"
      },
      {
        "literal": "Lingkai Kong"
      },
      {
        "literal": "Bo Dai"
      },
      {
        "literal": "Chao Zhang"
      }
    ],
    "citation-key": "arxiv_2305_16653",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "c254a26ebb8e1bcdca70b7581c780aa6b0dd4e5f5ded85e456f7e8fb8b38fcc6",
        "submitted_at": "2023-05-26T05:52:27Z",
        "updated_at": "2023-05-26T05:52:27Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2305.16653v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2305.16653",
      "catalog_id": "arxiv:2305.16653",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2023-05-26",
      "identifiers": {
        "arxiv": "2305.16653"
      },
      "record_updated_at": "2023-05-26"
    },
    "genre": "preprint",
    "id": "arxiv:2305.16653",
    "issued": {
      "date-parts": [
        [
          2023,
          5,
          26
        ]
      ]
    },
    "note": "First published: 2023-05-26; Recorded arXiv version: v1; Recorded version updated: 2023-05-26T05:52:27Z.",
    "title": "AdaPlanner: Adaptive Planning from Feedback with Language Models",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2306.07863v3",
    "archive": "arXiv",
    "archive_location": "2306.07863",
    "author": [
      {
        "literal": "Longtao Zheng"
      },
      {
        "literal": "Rundong Wang"
      },
      {
        "literal": "Xinrun Wang"
      },
      {
        "literal": "Bo An"
      }
    ],
    "citation-key": "arxiv_2306_07863",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "8fdb5fed2b2fa0ecbf5910dd6813d8de7910af286a1e4e00cdebb734a82e78b0",
        "submitted_at": "2023-06-13T15:49:41Z",
        "updated_at": "2024-01-19T06:59:26Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2306.07863v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2306.07863",
      "catalog_id": "arxiv:2306.07863",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2023-06-13",
      "identifiers": {
        "arxiv": "2306.07863"
      },
      "record_updated_at": "2024-01-19"
    },
    "genre": "preprint",
    "id": "arxiv:2306.07863",
    "issued": {
      "date-parts": [
        [
          2023,
          6,
          13
        ]
      ]
    },
    "note": "First published: 2023-06-13; Recorded arXiv version: v3; Recorded version updated: 2024-01-19T06:59:26Z.",
    "title": "Synapse: Trajectory-as-Exemplar Prompting with Memory for Computer Control",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2307.12856v4",
    "archive": "arXiv",
    "archive_location": "2307.12856",
    "author": [
      {
        "literal": "Izzeddin Gur"
      },
      {
        "literal": "Hiroki Furuta"
      },
      {
        "literal": "Austin Huang"
      },
      {
        "literal": "Mustafa Safdari"
      },
      {
        "literal": "Yutaka Matsuo"
      },
      {
        "literal": "Douglas Eck"
      },
      {
        "literal": "Aleksandra Faust"
      }
    ],
    "citation-key": "arxiv_2307_12856",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "65328d525a6458ef8c401fb0ac4a9b6a5142bad462bd15a212d124c83a56209f",
        "submitted_at": "2023-07-24T14:56:30Z",
        "updated_at": "2024-02-25T16:17:43Z",
        "version": 4,
        "versioned_url": "https://arxiv.org/abs/2307.12856v4"
      },
      "canonical_source_url": "https://arxiv.org/abs/2307.12856",
      "catalog_id": "arxiv:2307.12856",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2023-07-24",
      "identifiers": {
        "arxiv": "2307.12856"
      },
      "record_updated_at": "2024-02-25"
    },
    "genre": "preprint",
    "id": "arxiv:2307.12856",
    "issued": {
      "date-parts": [
        [
          2023,
          7,
          24
        ]
      ]
    },
    "note": "First published: 2023-07-24; Recorded arXiv version: v4; Recorded version updated: 2024-02-25T16:17:43Z.",
    "title": "A Real-World WebAgent with Planning, Long Context Understanding, and Program Synthesis",
    "type": "article",
    "version": "4"
  },
  {
    "URL": "https://arxiv.org/abs/2308.03688v3",
    "archive": "arXiv",
    "archive_location": "2308.03688",
    "author": [
      {
        "literal": "Xiao Liu"
      },
      {
        "literal": "Hao Yu"
      },
      {
        "literal": "Hanchen Zhang"
      },
      {
        "literal": "Yifan Xu"
      },
      {
        "literal": "Xuanyu Lei"
      },
      {
        "literal": "Hanyu Lai"
      },
      {
        "literal": "Yu Gu"
      },
      {
        "literal": "Hangliang Ding"
      },
      {
        "literal": "Kaiwen Men"
      },
      {
        "literal": "Kejuan Yang"
      },
      {
        "literal": "Shudan Zhang"
      },
      {
        "literal": "Xiang Deng"
      },
      {
        "literal": "Aohan Zeng"
      },
      {
        "literal": "Zhengxiao Du"
      },
      {
        "literal": "Chenhui Zhang"
      },
      {
        "literal": "Sheng Shen"
      },
      {
        "literal": "Tianjun Zhang"
      },
      {
        "literal": "Yu Su"
      },
      {
        "literal": "Huan Sun"
      },
      {
        "literal": "Minlie Huang"
      },
      {
        "literal": "Yuxiao Dong"
      },
      {
        "literal": "Jie Tang"
      }
    ],
    "citation-key": "arxiv_2308_03688",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d8c4ba4db4aeed3a7b7ce7f77d610d7e34b0e8db26a36463e2c0bdffcad8f22a",
        "submitted_at": "2023-08-07T16:08:11Z",
        "updated_at": "2025-10-04T03:54:18Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2308.03688v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2308.03688",
      "catalog_id": "arxiv:2308.03688",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2023-08-07",
      "identifiers": {
        "arxiv": "2308.03688"
      },
      "record_updated_at": "2025-10-04"
    },
    "genre": "preprint",
    "id": "arxiv:2308.03688",
    "issued": {
      "date-parts": [
        [
          2023,
          8,
          7
        ]
      ]
    },
    "note": "First published: 2023-08-07; Recorded arXiv version: v3; Recorded version updated: 2025-10-04T03:54:18Z.",
    "title": "AgentBench: Evaluating LLMs as Agents",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2309.08172v2",
    "archive": "arXiv",
    "archive_location": "2309.08172",
    "author": [
      {
        "literal": "Kaixin Ma"
      },
      {
        "literal": "Hongming Zhang"
      },
      {
        "literal": "Hongwei Wang"
      },
      {
        "literal": "Xiaoman Pan"
      },
      {
        "literal": "Wenhao Yu"
      },
      {
        "literal": "Dong Yu"
      }
    ],
    "citation-key": "arxiv_2309_08172",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "399c97382cac325a9d60a073a14221dee6ee6254cb0b8aa88a5341a7bb758c97",
        "submitted_at": "2023-09-15T05:44:08Z",
        "updated_at": "2024-02-21T17:42:32Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2309.08172v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2309.08172",
      "catalog_id": "arxiv:2309.08172",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2023-09-15",
      "identifiers": {
        "arxiv": "2309.08172"
      },
      "record_updated_at": "2024-02-21"
    },
    "genre": "preprint",
    "id": "arxiv:2309.08172",
    "issued": {
      "date-parts": [
        [
          2023,
          9,
          15
        ]
      ]
    },
    "note": "First published: 2023-09-15; Recorded arXiv version: v2; Recorded version updated: 2024-02-21T17:42:32Z.",
    "title": "LASER: LLM Agent with State-Space Exploration for Web Navigation",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2309.11436v4",
    "archive": "arXiv",
    "archive_location": "2309.11436",
    "author": [
      {
        "literal": "Zhuosheng Zhang"
      },
      {
        "literal": "Aston Zhang"
      }
    ],
    "citation-key": "arxiv_2309_11436",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "6d9cff1ece6115e87b7a9032f5720deb49694fbf570bea1f152018c4fa80d981",
        "submitted_at": "2023-09-20T16:12:32Z",
        "updated_at": "2024-06-07T04:52:29Z",
        "version": 4,
        "versioned_url": "https://arxiv.org/abs/2309.11436v4"
      },
      "canonical_source_url": "https://arxiv.org/abs/2309.11436",
      "catalog_id": "arxiv:2309.11436",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2023-09-20",
      "identifiers": {
        "arxiv": "2309.11436"
      },
      "publication": {
        "authors": [
          "Zhuosheng Zhang",
          "Aston Zhang"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2024/8. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2024,
          8
        ],
        "doi": "10.18653/v1/2024.findings-acl.186",
        "event_dates": null,
        "evidence_urls": [
          "https://aclanthology.org/2024.findings-acl.186/"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot.",
        "pages": "3132-3149",
        "pdf_url": "https://aclanthology.org/2024.findings-acl.186.pdf",
        "publisher": "Association for Computational Linguistics",
        "series": null,
        "title": "You Only Look at Screens: Multimodal Chain-of-Action Agents",
        "type": "conference-paper",
        "url": "https://aclanthology.org/2024.findings-acl.186/",
        "venue": "Findings of the Association for Computational Linguistics: ACL 2024",
        "volume": null
      },
      "record_updated_at": "2024-06-07"
    },
    "genre": "preprint",
    "id": "arxiv:2309.11436",
    "issued": {
      "date-parts": [
        [
          2023,
          9,
          20
        ]
      ]
    },
    "note": "First published: 2023-09-20; Recorded arXiv version: v4; Recorded version updated: 2024-06-07T04:52:29Z.",
    "title": "You Only Look at Screens: Multimodal Chain-of-Action Agents",
    "type": "article",
    "version": "4"
  },
  {
    "URL": "https://arxiv.org/abs/2310.03720v4",
    "archive": "arXiv",
    "archive_location": "2310.03720",
    "author": [
      {
        "literal": "Paloma Sodhi"
      },
      {
        "literal": "S. R. K. Branavan"
      },
      {
        "literal": "Yoav Artzi"
      },
      {
        "literal": "Ryan McDonald"
      }
    ],
    "citation-key": "arxiv_2310_03720",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d843f3c65e5961da773ec0265c2cfbf51d7c8214d9232ecad1638cc0dbb9ef64",
        "submitted_at": "2023-10-05T17:40:09Z",
        "updated_at": "2024-08-08T18:00:48Z",
        "version": 4,
        "versioned_url": "https://arxiv.org/abs/2310.03720v4"
      },
      "canonical_source_url": "https://arxiv.org/abs/2310.03720",
      "catalog_id": "arxiv:2310.03720",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2023-10-05",
      "identifiers": {
        "arxiv": "2310.03720"
      },
      "record_updated_at": "2024-08-08"
    },
    "genre": "preprint",
    "id": "arxiv:2310.03720",
    "issued": {
      "date-parts": [
        [
          2023,
          10,
          5
        ]
      ]
    },
    "note": "First published: 2023-10-05; Recorded arXiv version: v4; Recorded version updated: 2024-08-08T18:00:48Z.",
    "title": "SteP: Stacked LLM Policies for Web Actions",
    "type": "article",
    "version": "4"
  },
  {
    "URL": "https://arxiv.org/abs/2310.04869v1",
    "archive": "arXiv",
    "archive_location": "2310.04869",
    "author": [
      {
        "literal": "Yue Jiang"
      },
      {
        "literal": "Eldon Schoop"
      },
      {
        "literal": "Amanda Swearngin"
      },
      {
        "literal": "Jeffrey Nichols"
      }
    ],
    "citation-key": "arxiv_2310_04869",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "a49e879891017071fce4df0ff4b07bd9e8b5a6e14385af565549aadbde1bc1cf",
        "submitted_at": "2023-10-07T16:32:34Z",
        "updated_at": "2023-10-07T16:32:34Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2310.04869v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2310.04869",
      "catalog_id": "arxiv:2310.04869",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2023-10-07",
      "identifiers": {
        "arxiv": "2310.04869"
      },
      "record_updated_at": "2023-10-07"
    },
    "genre": "preprint",
    "id": "arxiv:2310.04869",
    "issued": {
      "date-parts": [
        [
          2023,
          10,
          7
        ]
      ]
    },
    "note": "First published: 2023-10-07; Recorded arXiv version: v1; Recorded version updated: 2023-10-07T16:32:34Z.",
    "title": "ILuvUI: Instruction-tuned LangUage-Vision modeling of UIs from Machine Conversations",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2310.08740v3",
    "archive": "arXiv",
    "archive_location": "2310.08740",
    "author": [
      {
        "literal": "Tao Li"
      },
      {
        "literal": "Gang Li"
      },
      {
        "literal": "Zhiwei Deng"
      },
      {
        "literal": "Bryan Wang"
      },
      {
        "literal": "Yang Li"
      }
    ],
    "citation-key": "arxiv_2310_08740",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "95092986184a232af62c61fa2aedefb837851dce9e0ff7a1128a67d47edfde40",
        "submitted_at": "2023-10-12T21:53:37Z",
        "updated_at": "2023-10-23T17:39:51Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2310.08740v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2310.08740",
      "catalog_id": "arxiv:2310.08740",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2023-10-12",
      "identifiers": {
        "arxiv": "2310.08740"
      },
      "record_updated_at": "2023-10-23"
    },
    "genre": "preprint",
    "id": "arxiv:2310.08740",
    "issued": {
      "date-parts": [
        [
          2023,
          10,
          12
        ]
      ]
    },
    "note": "First published: 2023-10-12; Recorded arXiv version: v3; Recorded version updated: 2023-10-23T17:39:51Z.",
    "title": "A Zero-Shot Language Agent for Computer Control with Structured Reflection",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2310.16042v2",
    "archive": "arXiv",
    "archive_location": "2310.16042",
    "author": [
      {
        "literal": "Heyi Tao"
      },
      {
        "literal": "Sethuraman T"
      },
      {
        "literal": "Michal Shlapentokh-Rothman"
      },
      {
        "literal": "Derek Hoiem"
      }
    ],
    "citation-key": "arxiv_2310_16042",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "7defced3e3beb77a5a6513dbfdff3f102edd922907b523bd062522e54720fa5e",
        "submitted_at": "2023-10-24T17:57:03Z",
        "updated_at": "2023-10-25T03:54:11Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2310.16042v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2310.16042",
      "catalog_id": "arxiv:2310.16042",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2023-10-24",
      "identifiers": {
        "arxiv": "2310.16042"
      },
      "record_updated_at": "2023-10-25"
    },
    "genre": "preprint",
    "id": "arxiv:2310.16042",
    "issued": {
      "date-parts": [
        [
          2023,
          10,
          24
        ]
      ]
    },
    "note": "First published: 2023-10-24; Recorded arXiv version: v2; Recorded version updated: 2023-10-25T03:54:11Z.",
    "title": "WebWISE: Web Interface Control and Sequential Exploration with Large Language Models",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2312.03003v3",
    "archive": "arXiv",
    "archive_location": "2312.03003",
    "author": [
      {
        "literal": "Sunjae Lee"
      },
      {
        "literal": "Junyoung Choi"
      },
      {
        "literal": "Jungjae Lee"
      },
      {
        "literal": "Munim Hasan Wasi"
      },
      {
        "literal": "Hojun Choi"
      },
      {
        "literal": "Steven Y. Ko"
      },
      {
        "literal": "Sangeun Oh"
      },
      {
        "literal": "Insik Shin"
      }
    ],
    "citation-key": "arxiv_2312_03003",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "b9b6f10b17513f565a13e035802d8bd681b5b0d468ba000df1427a8d59591e2c",
        "submitted_at": "2023-12-04T06:13:35Z",
        "updated_at": "2024-10-16T08:15:53Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2312.03003v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2312.03003",
      "catalog_id": "arxiv:2312.03003",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2023-12-04",
      "identifiers": {
        "arxiv": "2312.03003"
      },
      "record_updated_at": "2024-10-16"
    },
    "genre": "preprint",
    "id": "arxiv:2312.03003",
    "issued": {
      "date-parts": [
        [
          2023,
          12,
          4
        ]
      ]
    },
    "note": "First published: 2023-12-04; Recorded arXiv version: v3; Recorded version updated: 2024-10-16T08:15:53Z.",
    "title": "Explore, Select, Derive, and Recall: Augmenting LLM with Human-like Memory for Mobile Task Automation",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2312.06677v1",
    "archive": "arXiv",
    "archive_location": "2312.06677",
    "author": [
      {
        "literal": "Yanchu Guan"
      },
      {
        "literal": "Dong Wang"
      },
      {
        "literal": "Zhixuan Chu"
      },
      {
        "literal": "Shiyu Wang"
      },
      {
        "literal": "Feiyue Ni"
      },
      {
        "literal": "Ruihua Song"
      },
      {
        "literal": "Longfei Li"
      },
      {
        "literal": "Jinjie Gu"
      },
      {
        "literal": "Chenyi Zhuang"
      }
    ],
    "citation-key": "arxiv_2312_06677",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "598c8c30348d454c638944c932dd55355b5d3ecda6eb6f3910a27fffcab9622a",
        "submitted_at": "2023-12-04T07:51:58Z",
        "updated_at": "2023-12-04T07:51:58Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2312.06677v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2312.06677",
      "catalog_id": "arxiv:2312.06677",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2023-12-04",
      "identifiers": {
        "arxiv": "2312.06677"
      },
      "record_updated_at": "2023-12-04"
    },
    "genre": "preprint",
    "id": "arxiv:2312.06677",
    "issued": {
      "date-parts": [
        [
          2023,
          12,
          4
        ]
      ]
    },
    "note": "First published: 2023-12-04; Recorded arXiv version: v1; Recorded version updated: 2023-12-04T07:51:58Z.",
    "title": "Intelligent Virtual Assistants with LLM-based Process Automation",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2312.11190v2",
    "archive": "arXiv",
    "archive_location": "2312.11190",
    "author": [
      {
        "literal": "Yunpeng Song"
      },
      {
        "literal": "Yiheng Bian"
      },
      {
        "literal": "Yongtao Tang"
      },
      {
        "literal": "Guiyu Ma"
      },
      {
        "literal": "Zhongmin Cai"
      }
    ],
    "citation-key": "arxiv_2312_11190",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "df1390eab35db18dedaf135ddbbc9fdfcacbc9a70c51ec2609cf331cc05e06bd",
        "submitted_at": "2023-12-18T13:35:46Z",
        "updated_at": "2024-07-29T14:52:38Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2312.11190v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2312.11190",
      "catalog_id": "arxiv:2312.11190",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2023-12-18",
      "identifiers": {
        "arxiv": "2312.11190"
      },
      "record_updated_at": "2024-07-29"
    },
    "genre": "preprint",
    "id": "arxiv:2312.11190",
    "issued": {
      "date-parts": [
        [
          2023,
          12,
          18
        ]
      ]
    },
    "note": "First published: 2023-12-18; Recorded arXiv version: v2; Recorded version updated: 2024-07-29T14:52:38Z.",
    "title": "VisionTasker: Mobile Task Automation Using Vision Based UI Understanding and LLM Task Planning",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2312.13108v2",
    "archive": "arXiv",
    "archive_location": "2312.13108",
    "author": [
      {
        "literal": "Difei Gao"
      },
      {
        "literal": "Lei Ji"
      },
      {
        "literal": "Zechen Bai"
      },
      {
        "literal": "Mingyu Ouyang"
      },
      {
        "literal": "Peiran Li"
      },
      {
        "literal": "Dongxing Mao"
      },
      {
        "literal": "Qinchen Wu"
      },
      {
        "literal": "Weichen Zhang"
      },
      {
        "literal": "Peiyi Wang"
      },
      {
        "literal": "Xiangwu Guo"
      },
      {
        "literal": "Hengxu Wang"
      },
      {
        "literal": "Luowei Zhou"
      },
      {
        "literal": "Mike Zheng Shou"
      }
    ],
    "citation-key": "arxiv_2312_13108",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "c771b076044c170728041a4f54662d21493adcc3742d40db08cb2f9bd0614655",
        "submitted_at": "2023-12-20T15:28:38Z",
        "updated_at": "2024-01-01T14:26:39Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2312.13108v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2312.13108",
      "catalog_id": "arxiv:2312.13108",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2023-12-20",
      "identifiers": {
        "arxiv": "2312.13108"
      },
      "record_updated_at": "2024-01-01"
    },
    "genre": "preprint",
    "id": "arxiv:2312.13108",
    "issued": {
      "date-parts": [
        [
          2023,
          12,
          20
        ]
      ]
    },
    "note": "First published: 2023-12-20; Recorded arXiv version: v2; Recorded version updated: 2024-01-01T14:26:39Z.",
    "title": "ASSISTGUI: Task-Oriented Desktop Graphical User Interface Automation",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2401.16158v2",
    "archive": "arXiv",
    "archive_location": "2401.16158",
    "author": [
      {
        "literal": "Junyang Wang"
      },
      {
        "literal": "Haiyang Xu"
      },
      {
        "literal": "Jiabo Ye"
      },
      {
        "literal": "Ming Yan"
      },
      {
        "literal": "Weizhou Shen"
      },
      {
        "literal": "Ji Zhang"
      },
      {
        "literal": "Fei Huang"
      },
      {
        "literal": "Jitao Sang"
      }
    ],
    "citation-key": "arxiv_2401_16158",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "81ac041a3e5f98bfc23c91f768b32ad7adc3f28d8c9d55a285bb29906c8b5a9c",
        "submitted_at": "2024-01-29T13:46:37Z",
        "updated_at": "2024-04-18T06:53:38Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2401.16158v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2401.16158",
      "catalog_id": "arxiv:2401.16158",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-01-29",
      "identifiers": {
        "arxiv": "2401.16158"
      },
      "record_updated_at": "2024-04-18"
    },
    "genre": "preprint",
    "id": "arxiv:2401.16158",
    "issued": {
      "date-parts": [
        [
          2024,
          1,
          29
        ]
      ]
    },
    "note": "First published: 2024-01-29; Recorded arXiv version: v2; Recorded version updated: 2024-04-18T06:53:38Z.",
    "title": "Mobile-Agent: Autonomous Multi-Modal Mobile Device Agent with Visual Perception",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2402.07456v2",
    "archive": "arXiv",
    "archive_location": "2402.07456",
    "author": [
      {
        "literal": "Zhiyong Wu"
      },
      {
        "literal": "Chengcheng Han"
      },
      {
        "literal": "Zichen Ding"
      },
      {
        "literal": "Zhenmin Weng"
      },
      {
        "literal": "Zhoumianze Liu"
      },
      {
        "literal": "Shunyu Yao"
      },
      {
        "literal": "Tao Yu"
      },
      {
        "literal": "Lingpeng Kong"
      }
    ],
    "citation-key": "arxiv_2402_07456",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "674b005f670a3593bd215a8ff75eac05c90ee068578fc26afdd998c40f5b7b71",
        "submitted_at": "2024-02-12T07:29:22Z",
        "updated_at": "2024-02-15T09:30:48Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2402.07456v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2402.07456",
      "catalog_id": "arxiv:2402.07456",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-02-12",
      "identifiers": {
        "arxiv": "2402.07456"
      },
      "record_updated_at": "2024-02-15"
    },
    "genre": "preprint",
    "id": "arxiv:2402.07456",
    "issued": {
      "date-parts": [
        [
          2024,
          2,
          12
        ]
      ]
    },
    "note": "First published: 2024-02-12; Recorded arXiv version: v2; Recorded version updated: 2024-02-15T09:30:48Z.",
    "title": "OS-Copilot: Towards Generalist Computer Agents with Self-Improvement",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2402.07939v5",
    "archive": "arXiv",
    "archive_location": "2402.07939",
    "author": [
      {
        "literal": "Chaoyun Zhang"
      },
      {
        "literal": "Liqun Li"
      },
      {
        "literal": "Shilin He"
      },
      {
        "literal": "Xu Zhang"
      },
      {
        "literal": "Bo Qiao"
      },
      {
        "literal": "Si Qin"
      },
      {
        "literal": "Minghua Ma"
      },
      {
        "literal": "Yu Kang"
      },
      {
        "literal": "Qingwei Lin"
      },
      {
        "literal": "Saravan Rajmohan"
      },
      {
        "literal": "Dongmei Zhang"
      },
      {
        "literal": "Qi Zhang"
      }
    ],
    "citation-key": "arxiv_2402_07939",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "533acb1c9aa3f1c77adb2c67779da86e3d5de95264ba3ec1f81d803f67da7753",
        "submitted_at": "2024-02-08T15:40:35Z",
        "updated_at": "2024-05-23T05:18:58Z",
        "version": 5,
        "versioned_url": "https://arxiv.org/abs/2402.07939v5"
      },
      "canonical_source_url": "https://arxiv.org/abs/2402.07939",
      "catalog_id": "arxiv:2402.07939",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2024-02-08",
      "identifiers": {
        "arxiv": "2402.07939"
      },
      "publication": {
        "authors": [
          "Chaoyun Zhang",
          "Liqun Li",
          "Shilin He",
          "Xu Zhang",
          "Bo Qiao",
          "Si Qin",
          "Minghua Ma",
          "Yu Kang",
          "Qingwei Lin",
          "Saravan Rajmohan",
          "Dongmei Zhang",
          "Qi Zhang"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2025/4. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2025,
          4
        ],
        "doi": "10.18653/v1/2025.naacl-long.26",
        "event_dates": null,
        "evidence_urls": [
          "https://aclanthology.org/2025.naacl-long.26/"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot.",
        "pages": "597-622",
        "pdf_url": "https://aclanthology.org/2025.naacl-long.26.pdf",
        "publisher": "Association for Computational Linguistics",
        "series": null,
        "title": "UFO: A UI-Focused Agent for Windows OS Interaction",
        "type": "conference-paper",
        "url": "https://aclanthology.org/2025.naacl-long.26/",
        "venue": "Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)",
        "volume": null
      },
      "record_updated_at": "2024-05-23"
    },
    "genre": "preprint",
    "id": "arxiv:2402.07939",
    "issued": {
      "date-parts": [
        [
          2024,
          2,
          8
        ]
      ]
    },
    "note": "First published: 2024-02-08; Recorded arXiv version: v5; Recorded version updated: 2024-05-23T05:18:58Z.",
    "title": "UFO: A UI-Focused Agent for Windows OS Interaction",
    "type": "article",
    "version": "5"
  },
  {
    "URL": "https://arxiv.org/abs/2402.07945v1",
    "archive": "arXiv",
    "archive_location": "2402.07945",
    "author": [
      {
        "literal": "Runliang Niu"
      },
      {
        "literal": "Jindong Li"
      },
      {
        "literal": "Shiqi Wang"
      },
      {
        "literal": "Yali Fu"
      },
      {
        "literal": "Xiyu Hu"
      },
      {
        "literal": "Xueyuan Leng"
      },
      {
        "literal": "He Kong"
      },
      {
        "literal": "Yi Chang"
      },
      {
        "literal": "Qi Wang"
      }
    ],
    "citation-key": "arxiv_2402_07945",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "63424a86302625b15a91ff0252ca16c0a6448160bfd748ba32e386c82883584e",
        "submitted_at": "2024-02-09T02:33:45Z",
        "updated_at": "2024-02-09T02:33:45Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2402.07945v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2402.07945",
      "catalog_id": "arxiv:2402.07945",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-02-09",
      "identifiers": {
        "arxiv": "2402.07945"
      },
      "record_updated_at": "2024-02-09"
    },
    "genre": "preprint",
    "id": "arxiv:2402.07945",
    "issued": {
      "date-parts": [
        [
          2024,
          2,
          9
        ]
      ]
    },
    "note": "First published: 2024-02-09; Recorded arXiv version: v1; Recorded version updated: 2024-02-09T02:33:45Z.",
    "title": "ScreenAgent: A Vision Language Model-driven Computer Control Agent",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2403.02713v2",
    "archive": "arXiv",
    "archive_location": "2403.02713",
    "author": [
      {
        "literal": "Jiwen Zhang"
      },
      {
        "literal": "Jihao Wu"
      },
      {
        "literal": "Yihua Teng"
      },
      {
        "literal": "Minghui Liao"
      },
      {
        "literal": "Nuo Xu"
      },
      {
        "literal": "Xiao Xiao"
      },
      {
        "literal": "Zhongyu Wei"
      },
      {
        "literal": "Duyu Tang"
      }
    ],
    "citation-key": "arxiv_2403_02713",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "4de9e7d68aa8450cab65ecb300524e85b1d2b2a91d58fecef370deceaba64b1e",
        "submitted_at": "2024-03-05T07:09:35Z",
        "updated_at": "2024-07-13T02:12:30Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2403.02713v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2403.02713",
      "catalog_id": "arxiv:2403.02713",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2024-03-05",
      "identifiers": {
        "arxiv": "2403.02713"
      },
      "publication": {
        "authors": [
          "Jiwen Zhang (张霁雯)",
          "Jihao Wu",
          "Teng Yihua",
          "Minghui Liao",
          "Nuo Xu",
          "Xiao Xiao",
          "Zhongyu Wei (魏忠钰)",
          "Duyu Tang"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2024/11. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2024,
          11
        ],
        "doi": "10.18653/v1/2024.findings-emnlp.702",
        "event_dates": null,
        "evidence_urls": [
          "https://aclanthology.org/2024.findings-emnlp.702/"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity. The normalized publisher abstract matches the retained arXiv abstract SHA-256.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot. Publisher spelling, initials, name annotations or author lists differ from the retained preprint.",
        "pages": "12016-12031",
        "pdf_url": "https://aclanthology.org/2024.findings-emnlp.702.pdf",
        "publisher": "Association for Computational Linguistics",
        "series": null,
        "title": "Android in the Zoo: Chain-of-Action-Thought for GUI Agents",
        "type": "conference-paper",
        "url": "https://aclanthology.org/2024.findings-emnlp.702/",
        "venue": "Findings of the Association for Computational Linguistics: EMNLP 2024",
        "volume": null
      },
      "record_updated_at": "2024-07-13"
    },
    "genre": "preprint",
    "id": "arxiv:2403.02713",
    "issued": {
      "date-parts": [
        [
          2024,
          3,
          5
        ]
      ]
    },
    "note": "First published: 2024-03-05; Recorded arXiv version: v2; Recorded version updated: 2024-07-13T02:12:30Z.",
    "title": "Android in the Zoo: Chain-of-Action-Thought for GUI Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2403.03186v3",
    "archive": "arXiv",
    "archive_location": "2403.03186",
    "author": [
      {
        "literal": "Weihao Tan"
      },
      {
        "literal": "Wentao Zhang"
      },
      {
        "literal": "Xinrun Xu"
      },
      {
        "literal": "Haochong Xia"
      },
      {
        "literal": "Ziluo Ding"
      },
      {
        "literal": "Boyu Li"
      },
      {
        "literal": "Bohan Zhou"
      },
      {
        "literal": "Junpeng Yue"
      },
      {
        "literal": "Jiechuan Jiang"
      },
      {
        "literal": "Yewen Li"
      },
      {
        "literal": "Ruyi An"
      },
      {
        "literal": "Molei Qin"
      },
      {
        "literal": "Chuqiao Zong"
      },
      {
        "literal": "Longtao Zheng"
      },
      {
        "literal": "Yujie Wu"
      },
      {
        "literal": "Xiaoqiang Chai"
      },
      {
        "literal": "Yifei Bi"
      },
      {
        "literal": "Tianbao Xie"
      },
      {
        "literal": "Pengjie Gu"
      },
      {
        "literal": "Xiyun Li"
      },
      {
        "literal": "Ceyao Zhang"
      },
      {
        "literal": "Long Tian"
      },
      {
        "literal": "Chaojie Wang"
      },
      {
        "literal": "Xinrun Wang"
      },
      {
        "literal": "Börje F. Karlsson"
      },
      {
        "literal": "Bo An"
      },
      {
        "literal": "Shuicheng Yan"
      },
      {
        "literal": "Zongqing Lu"
      }
    ],
    "citation-key": "arxiv_2403_03186",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "63c68c596f1c81485e77f98beae696deca9785a5d8725ab4456ee35870e64a99",
        "submitted_at": "2024-03-05T18:22:29Z",
        "updated_at": "2024-07-02T17:23:13Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2403.03186v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2403.03186",
      "catalog_id": "arxiv:2403.03186",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2024-03-05",
      "identifiers": {
        "arxiv": "2403.03186"
      },
      "publication": {
        "authors": [
          "Weihao Tan",
          "Wentao Zhang",
          "Xinrun Xu",
          "Haochong Xia",
          "Ziluo Ding",
          "Boyu Li",
          "Bohan Zhou",
          "Junpeng Yue",
          "Jiechuan Jiang",
          "Yewen Li",
          "Ruyi An",
          "Molei Qin",
          "Chuqiao Zong",
          "Longtao Zheng",
          "Yujie Wu",
          "Xiaoqiang Chai",
          "Yifei Bi",
          "Tianbao Xie",
          "Pengjie Gu",
          "Xiyun Li",
          "Ceyao Zhang",
          "Long Tian",
          "Chaojie Wang",
          "Xinrun Wang",
          "Börje F. Karlsson",
          "Bo An",
          "Shuicheng Yan",
          "Zongqing Lu"
        ],
        "date_basis": "Explicit publisher metadata: 2025/10/06; precision=day.",
        "date_parts": [
          2025,
          10,
          6
        ],
        "doi": null,
        "event_dates": "13--19 Jul 2025",
        "evidence_urls": [
          "https://proceedings.mlr.press/v267/tan25h.html"
        ],
        "identity_basis": "Same work: exact normalized landing title, corresponding author list, and matching specific contribution. Publisher and arXiv versions remain distinct.",
        "issue": null,
        "notes": "The publisher issue date comes from citation_publication_date; the conference date range is retained separately and does not replace it. Publisher-hosted proceedings record; distinct from the retained arXiv preprint snapshot.",
        "pages": "58658-58725",
        "pdf_url": "https://raw.githubusercontent.com/mlresearch/v267/main/assets/tan25h/tan25h.pdf",
        "publisher": "PMLR",
        "series": "Proceedings of Machine Learning Research",
        "title": "Cradle: Empowering Foundation Agents towards General Computer Control",
        "type": "conference-paper",
        "url": "https://proceedings.mlr.press/v267/tan25h.html",
        "venue": "Proceedings of the 42nd International Conference on Machine Learning",
        "volume": "267"
      },
      "record_updated_at": "2024-07-02"
    },
    "genre": "preprint",
    "id": "arxiv:2403.03186",
    "issued": {
      "date-parts": [
        [
          2024,
          3,
          5
        ]
      ]
    },
    "note": "First published: 2024-03-05; Recorded arXiv version: v3; Recorded version updated: 2024-07-02T17:23:13Z.",
    "title": "Cradle: Empowering Foundation Agents Towards General Computer Control",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2404.18074v3",
    "archive": "arXiv",
    "archive_location": "2404.18074",
    "author": [
      {
        "literal": "Zirui Song"
      },
      {
        "literal": "Yaohang Li"
      },
      {
        "literal": "Meng Fang"
      },
      {
        "literal": "Yanda Li"
      },
      {
        "literal": "Zhenhao Chen"
      },
      {
        "literal": "Zecheng Shi"
      },
      {
        "literal": "Yuan Huang"
      },
      {
        "literal": "Xiuying Chen"
      },
      {
        "literal": "Ling Chen"
      }
    ],
    "citation-key": "arxiv_2404_18074",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "5e531ec2ef9e5d90586b8f3052e16d41da95114b7382047d0a02fd671a0db355",
        "submitted_at": "2024-04-28T05:33:15Z",
        "updated_at": "2025-03-23T13:04:57Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2404.18074v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2404.18074",
      "catalog_id": "arxiv:2404.18074",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-04-28",
      "identifiers": {
        "arxiv": "2404.18074"
      },
      "record_updated_at": "2025-03-23"
    },
    "genre": "preprint",
    "id": "arxiv:2404.18074",
    "issued": {
      "date-parts": [
        [
          2024,
          4,
          28
        ]
      ]
    },
    "note": "First published: 2024-04-28; Recorded arXiv version: v3; Recorded version updated: 2025-03-23T13:04:57Z.",
    "title": "MMAC-Copilot: Multi-modal Agent Collaboration Operating Copilot",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2405.04497v2",
    "archive": "arXiv",
    "archive_location": "2405.04497",
    "author": [
      {
        "literal": "Kihoon Son"
      },
      {
        "literal": "Jinhyeon Kwon"
      },
      {
        "literal": "DaEun Choi"
      },
      {
        "literal": "Tae Soo Kim"
      },
      {
        "literal": "Young-Ho Kim"
      },
      {
        "literal": "Sangdoo Yun"
      },
      {
        "literal": "Juho Kim"
      }
    ],
    "citation-key": "arxiv_2405_04497",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "a358d417d8bef89a01e799d88564e3ede4678ebda4950e87c96cf1e1954b85e1",
        "submitted_at": "2024-05-07T17:10:31Z",
        "updated_at": "2024-05-08T05:44:21Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2405.04497v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2405.04497",
      "catalog_id": "arxiv:2405.04497",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-05-07",
      "identifiers": {
        "arxiv": "2405.04497"
      },
      "record_updated_at": "2024-05-08"
    },
    "genre": "preprint",
    "id": "arxiv:2405.04497",
    "issued": {
      "date-parts": [
        [
          2024,
          5,
          7
        ]
      ]
    },
    "note": "First published: 2024-05-07; Recorded arXiv version: v2; Recorded version updated: 2024-05-08T05:44:21Z.",
    "title": "Unveiling Disparities in Web Task Handling Between Human and Web Agent",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2405.11120v1",
    "archive": "arXiv",
    "archive_location": "2405.11120",
    "author": [
      {
        "literal": "William E Bishop"
      },
      {
        "literal": "Alice Li"
      },
      {
        "literal": "Christopher Rawles"
      },
      {
        "literal": "Oriana Riva"
      }
    ],
    "citation-key": "arxiv_2405_11120",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "81f62326dddc5c512fb05472eb1f387064c0a5aad77453e9d2e85415bb627fa9",
        "submitted_at": "2024-05-17T23:27:33Z",
        "updated_at": "2024-05-17T23:27:33Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2405.11120v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2405.11120",
      "catalog_id": "arxiv:2405.11120",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-05-17",
      "identifiers": {
        "arxiv": "2405.11120"
      },
      "record_updated_at": "2024-05-17"
    },
    "genre": "preprint",
    "id": "arxiv:2405.11120",
    "issued": {
      "date-parts": [
        [
          2024,
          5,
          17
        ]
      ]
    },
    "note": "First published: 2024-05-17; Recorded arXiv version: v1; Recorded version updated: 2024-05-17T23:27:33Z.",
    "title": "Latent State Estimation Helps UI Agents to Reason",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2406.01014v1",
    "archive": "arXiv",
    "archive_location": "2406.01014",
    "author": [
      {
        "literal": "Junyang Wang"
      },
      {
        "literal": "Haiyang Xu"
      },
      {
        "literal": "Haitao Jia"
      },
      {
        "literal": "Xi Zhang"
      },
      {
        "literal": "Ming Yan"
      },
      {
        "literal": "Weizhou Shen"
      },
      {
        "literal": "Ji Zhang"
      },
      {
        "literal": "Fei Huang"
      },
      {
        "literal": "Jitao Sang"
      }
    ],
    "citation-key": "arxiv_2406_01014",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "a1338f3543a04bc2f9734e5fcabbfd5ebda6ac263461c2305f7c39383e938573",
        "submitted_at": "2024-06-03T05:50:00Z",
        "updated_at": "2024-06-03T05:50:00Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2406.01014v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2406.01014",
      "catalog_id": "arxiv:2406.01014",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-06-03",
      "identifiers": {
        "arxiv": "2406.01014"
      },
      "record_updated_at": "2024-06-03"
    },
    "genre": "preprint",
    "id": "arxiv:2406.01014",
    "issued": {
      "date-parts": [
        [
          2024,
          6,
          3
        ]
      ]
    },
    "note": "First published: 2024-06-03; Recorded arXiv version: v1; Recorded version updated: 2024-06-03T05:50:00Z.",
    "title": "Mobile-Agent-v2: Mobile Device Operation Assistant with Effective Navigation via Multi-Agent Collaboration",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2406.06947v3",
    "archive": "arXiv",
    "archive_location": "2406.06947",
    "author": [
      {
        "literal": "Junhee Cho"
      },
      {
        "literal": "Jihoon Kim"
      },
      {
        "literal": "Daseul Bae"
      },
      {
        "literal": "Jinho Choo"
      },
      {
        "literal": "Youngjune Gwon"
      },
      {
        "literal": "Yeong-Dae Kwon"
      }
    ],
    "citation-key": "arxiv_2406_06947",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "5f85b2cf7650f41ba9ee6e4ae83cd9041e077a44adcb685e702d644bd5779963",
        "submitted_at": "2024-06-11T05:21:20Z",
        "updated_at": "2024-12-26T11:02:28Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2406.06947v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2406.06947",
      "catalog_id": "arxiv:2406.06947",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-06-11",
      "identifiers": {
        "arxiv": "2406.06947"
      },
      "record_updated_at": "2024-12-26"
    },
    "genre": "preprint",
    "id": "arxiv:2406.06947",
    "issued": {
      "date-parts": [
        [
          2024,
          6,
          11
        ]
      ]
    },
    "note": "First published: 2024-06-11; Recorded arXiv version: v3; Recorded version updated: 2024-12-26T11:02:28Z.",
    "title": "CAAP: Context-Aware Action Planning Prompting to Solve Computer Tasks with Front-End UI Only",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2406.10227v1",
    "archive": "arXiv",
    "archive_location": "2406.10227",
    "author": [
      {
        "literal": "Kevin Qinghong Lin"
      },
      {
        "literal": "Linjie Li"
      },
      {
        "literal": "Difei Gao"
      },
      {
        "literal": "Qinchen WU"
      },
      {
        "literal": "Mingyi Yan"
      },
      {
        "literal": "Zhengyuan Yang"
      },
      {
        "literal": "Lijuan Wang"
      },
      {
        "literal": "Mike Zheng Shou"
      }
    ],
    "citation-key": "arxiv_2406_10227",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "66b5fb1e839490df80e9750fdc049f7fe3d2d4686c4090d78ea01bc542fe55b0",
        "submitted_at": "2024-06-14T17:59:08Z",
        "updated_at": "2024-06-14T17:59:08Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2406.10227v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2406.10227",
      "catalog_id": "arxiv:2406.10227",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2024-06-14",
      "identifiers": {
        "arxiv": "2406.10227"
      },
      "publication": {
        "authors": [
          "Kevin Q. Lin",
          "Linjie Li",
          "Difei Gao",
          "Qinchen Wu",
          "Mingyi Yan",
          "Zhengyuan Yang",
          "Lijuan Wang",
          "Mike Z. Shou"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2024-12-16. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2024,
          12,
          16
        ],
        "doi": "10.52202/079017-2214",
        "event_dates": null,
        "evidence_urls": [
          "https://proceedings.neurips.cc/paper_files/paper/2024/hash/804e757b7d7043c26701c3a313032101-Abstract-Datasets_and_Benchmarks_Track.html"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot. Publisher spelling, initials, name annotations or author lists differ from the retained preprint.",
        "pages": "69329-69360",
        "pdf_url": "https://proceedings.neurips.cc/paper_files/paper/2024/file/804e757b7d7043c26701c3a313032101-Paper-Datasets_and_Benchmarks_Track.pdf",
        "publisher": null,
        "series": null,
        "title": "VideoGUI: A Benchmark for GUI Automation from Instructional Videos",
        "type": "conference-paper",
        "url": "https://proceedings.neurips.cc/paper_files/paper/2024/hash/804e757b7d7043c26701c3a313032101-Abstract-Datasets_and_Benchmarks_Track.html",
        "venue": "Advances in Neural Information Processing Systems",
        "volume": "37"
      },
      "record_updated_at": "2024-06-14"
    },
    "genre": "preprint",
    "id": "arxiv:2406.10227",
    "issued": {
      "date-parts": [
        [
          2024,
          6,
          14
        ]
      ]
    },
    "note": "First published: 2024-06-14; Recorded arXiv version: v1; Recorded version updated: 2024-06-14T17:59:08Z.",
    "title": "VideoGUI: A Benchmark for GUI Automation from Instructional Videos",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2407.01476v4",
    "archive": "arXiv",
    "archive_location": "2407.01476",
    "author": [
      {
        "literal": "Jing Yu Koh"
      },
      {
        "literal": "Stephen McAleer"
      },
      {
        "literal": "Daniel Fried"
      },
      {
        "literal": "Ruslan Salakhutdinov"
      }
    ],
    "citation-key": "arxiv_2407_01476",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "ec7aaf2a34e38f72da652bb2d8350c951229c154e2bc368a927a79287b36c439",
        "submitted_at": "2024-07-01T17:07:55Z",
        "updated_at": "2026-02-08T15:06:40Z",
        "version": 4,
        "versioned_url": "https://arxiv.org/abs/2407.01476v4"
      },
      "canonical_source_url": "https://arxiv.org/abs/2407.01476",
      "catalog_id": "arxiv:2407.01476",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-07-01",
      "identifiers": {
        "arxiv": "2407.01476"
      },
      "record_updated_at": "2026-02-08"
    },
    "genre": "preprint",
    "id": "arxiv:2407.01476",
    "issued": {
      "date-parts": [
        [
          2024,
          7,
          1
        ]
      ]
    },
    "note": "First published: 2024-07-01; Recorded arXiv version: v4; Recorded version updated: 2026-02-08T15:06:40Z.",
    "title": "Tree Search for Language Model Agents",
    "type": "article",
    "version": "4"
  },
  {
    "URL": "https://arxiv.org/abs/2407.03913v1",
    "archive": "arXiv",
    "archive_location": "2407.03913",
    "author": [
      {
        "literal": "Jiayi Zhang"
      },
      {
        "literal": "Chuang Zhao"
      },
      {
        "literal": "Yihan Zhao"
      },
      {
        "literal": "Zhaoyang Yu"
      },
      {
        "literal": "Ming He"
      },
      {
        "literal": "Jianping Fan"
      }
    ],
    "citation-key": "arxiv_2407_03913",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "aa74755c53e20b3a9ff4b72e03a557f455ecceaefc96a50587aefd78ac3c8176",
        "submitted_at": "2024-07-04T13:12:19Z",
        "updated_at": "2024-07-04T13:12:19Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2407.03913v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2407.03913",
      "catalog_id": "arxiv:2407.03913",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-07-04",
      "identifiers": {
        "arxiv": "2407.03913"
      },
      "record_updated_at": "2024-07-04"
    },
    "genre": "preprint",
    "id": "arxiv:2407.03913",
    "issued": {
      "date-parts": [
        [
          2024,
          7,
          4
        ]
      ]
    },
    "note": "First published: 2024-07-04; Recorded arXiv version: v1; Recorded version updated: 2024-07-04T13:12:19Z.",
    "title": "MobileExperts: A Dynamic Tool-Enabled Agent Team in Mobile Devices",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2407.05291v2",
    "archive": "arXiv",
    "archive_location": "2407.05291",
    "author": [
      {
        "literal": "Léo Boisvert"
      },
      {
        "literal": "Megh Thakkar"
      },
      {
        "literal": "Maxime Gasse"
      },
      {
        "literal": "Massimo Caccia"
      },
      {
        "literal": "Thibault Le Sellier De Chezelles"
      },
      {
        "literal": "Quentin Cappart"
      },
      {
        "literal": "Nicolas Chapados"
      },
      {
        "literal": "Alexandre Lacoste"
      },
      {
        "literal": "Alexandre Drouin"
      }
    ],
    "citation-key": "arxiv_2407_05291",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "2b56a6208e4fa3523cae3fbae673affcd5fe3d6a9c19df482d27fe71d43ffef5",
        "submitted_at": "2024-07-07T07:15:49Z",
        "updated_at": "2025-02-05T21:50:07Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2407.05291v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2407.05291",
      "catalog_id": "arxiv:2407.05291",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2024-07-07",
      "identifiers": {
        "arxiv": "2407.05291"
      },
      "publication": {
        "authors": [
          "Léo Boisvert",
          "Megh Thakkar",
          "Maxime Gasse",
          "Massimo Caccia",
          "Thibault L. De Chezelles",
          "Quentin Cappart",
          "Nicolas Chapados",
          "Alexandre Lacoste",
          "Alexandre Drouin"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2024-12-16. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2024,
          12,
          16
        ],
        "doi": "10.52202/079017-0195",
        "event_dates": null,
        "evidence_urls": [
          "https://proceedings.neurips.cc/paper_files/paper/2024/hash/0b82662b6c32e887bb252a74d8cb2d5e-Abstract-Datasets_and_Benchmarks_Track.html"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot. Publisher spelling, initials, name annotations or author lists differ from the retained preprint.",
        "pages": "5996-6051",
        "pdf_url": "https://proceedings.neurips.cc/paper_files/paper/2024/file/0b82662b6c32e887bb252a74d8cb2d5e-Paper-Datasets_and_Benchmarks_Track.pdf",
        "publisher": null,
        "series": null,
        "title": "WorkArena++: Towards Compositional Planning and Reasoning-based Common Knowledge Work Tasks",
        "type": "conference-paper",
        "url": "https://proceedings.neurips.cc/paper_files/paper/2024/hash/0b82662b6c32e887bb252a74d8cb2d5e-Abstract-Datasets_and_Benchmarks_Track.html",
        "venue": "Advances in Neural Information Processing Systems",
        "volume": "37"
      },
      "record_updated_at": "2025-02-05"
    },
    "genre": "preprint",
    "id": "arxiv:2407.05291",
    "issued": {
      "date-parts": [
        [
          2024,
          7,
          7
        ]
      ]
    },
    "note": "First published: 2024-07-07; Recorded arXiv version: v2; Recorded version updated: 2025-02-05T21:50:07Z.",
    "title": "WorkArena++: Towards Compositional Planning and Reasoning-based Common Knowledge Work Tasks",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2407.13032v1",
    "archive": "arXiv",
    "archive_location": "2407.13032",
    "author": [
      {
        "literal": "Tamer Abuelsaad"
      },
      {
        "literal": "Deepak Akkil"
      },
      {
        "literal": "Prasenjit Dey"
      },
      {
        "literal": "Ashish Jagmohan"
      },
      {
        "literal": "Aditya Vempaty"
      },
      {
        "literal": "Ravi Kokku"
      }
    ],
    "citation-key": "arxiv_2407_13032",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "18cfd29fcb66d703d90f747c162db1b3fc6a43f400b4eacbddf081bb83ed7947",
        "submitted_at": "2024-07-17T21:44:28Z",
        "updated_at": "2024-07-17T21:44:28Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2407.13032v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2407.13032",
      "catalog_id": "arxiv:2407.13032",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-07-17",
      "identifiers": {
        "arxiv": "2407.13032"
      },
      "record_updated_at": "2024-07-17"
    },
    "genre": "preprint",
    "id": "arxiv:2407.13032",
    "issued": {
      "date-parts": [
        [
          2024,
          7,
          17
        ]
      ]
    },
    "note": "First published: 2024-07-17; Recorded arXiv version: v1; Recorded version updated: 2024-07-17T21:44:28Z.",
    "title": "Agent-E: From Autonomous Web Navigation to Foundational Design Principles in Agentic Systems",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2407.15711v2",
    "archive": "arXiv",
    "archive_location": "2407.15711",
    "author": [
      {
        "literal": "Ori Yoran"
      },
      {
        "literal": "Samuel Joseph Amouyal"
      },
      {
        "literal": "Chaitanya Malaviya"
      },
      {
        "literal": "Ben Bogin"
      },
      {
        "literal": "Ofir Press"
      },
      {
        "literal": "Jonathan Berant"
      }
    ],
    "citation-key": "arxiv_2407_15711",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "2f028fb7f4d64fd4c29886d49b5d6f0fabddba309b669e019edf13586deb9c90",
        "submitted_at": "2024-07-22T15:18:45Z",
        "updated_at": "2024-10-21T15:45:31Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2407.15711v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2407.15711",
      "catalog_id": "arxiv:2407.15711",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2024-07-22",
      "identifiers": {
        "arxiv": "2407.15711"
      },
      "publication": {
        "authors": [
          "Ori Yoran",
          "Samuel Joseph Amouyal",
          "Chaitanya Malaviya",
          "Ben Bogin",
          "Ofir Press",
          "Jonathan Berant"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2024/11. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2024,
          11
        ],
        "doi": "10.18653/v1/2024.emnlp-main.505",
        "event_dates": null,
        "evidence_urls": [
          "https://aclanthology.org/2024.emnlp-main.505/"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot.",
        "pages": "8938-8968",
        "pdf_url": "https://aclanthology.org/2024.emnlp-main.505.pdf",
        "publisher": "Association for Computational Linguistics",
        "series": null,
        "title": "AssistantBench: Can Web Agents Solve Realistic and Time-Consuming Tasks?",
        "type": "conference-paper",
        "url": "https://aclanthology.org/2024.emnlp-main.505/",
        "venue": "Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing",
        "volume": null
      },
      "record_updated_at": "2024-10-21"
    },
    "genre": "preprint",
    "id": "arxiv:2407.15711",
    "issued": {
      "date-parts": [
        [
          2024,
          7,
          22
        ]
      ]
    },
    "note": "First published: 2024-07-22; Recorded arXiv version: v2; Recorded version updated: 2024-10-21T15:45:31Z.",
    "title": "AssistantBench: Can Web Agents Solve Realistic and Time-Consuming Tasks?",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2408.15978v1",
    "archive": "arXiv",
    "archive_location": "2408.15978",
    "author": [
      {
        "literal": "Yao Zhang"
      },
      {
        "literal": "Zijian Ma"
      },
      {
        "literal": "Yunpu Ma"
      },
      {
        "literal": "Zhen Han"
      },
      {
        "literal": "Yu Wu"
      },
      {
        "literal": "Volker Tresp"
      }
    ],
    "citation-key": "arxiv_2408_15978",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "9e3a34f16101df7a96ffdabbdf86b060e195494f16499f20bdf54110a0394cae",
        "submitted_at": "2024-08-28T17:49:29Z",
        "updated_at": "2024-08-28T17:49:29Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2408.15978v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2408.15978",
      "catalog_id": "arxiv:2408.15978",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-08-28",
      "identifiers": {
        "arxiv": "2408.15978"
      },
      "record_updated_at": "2024-08-28"
    },
    "genre": "preprint",
    "id": "arxiv:2408.15978",
    "issued": {
      "date-parts": [
        [
          2024,
          8,
          28
        ]
      ]
    },
    "note": "First published: 2024-08-28; Recorded arXiv version: v1; Recorded version updated: 2024-08-28T17:49:29Z.",
    "title": "WebPilot: A Versatile and Autonomous Multi-Agent System for Web Task Execution with Strategic Exploration",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2409.01927v1",
    "archive": "arXiv",
    "archive_location": "2409.01927",
    "author": [
      {
        "literal": "Segev Shlomov"
      },
      {
        "literal": "Ben wiesel"
      },
      {
        "literal": "Aviad Sela"
      },
      {
        "literal": "Ido Levy"
      },
      {
        "literal": "Liane Galanti"
      },
      {
        "literal": "Roy Abitbol"
      }
    ],
    "citation-key": "arxiv_2409_01927",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "cdb228e652218d7f8981d9d16ca29b465004675c510d1e9c5c7b543bec9dc454",
        "submitted_at": "2024-09-03T14:17:09Z",
        "updated_at": "2024-09-03T14:17:09Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2409.01927v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2409.01927",
      "catalog_id": "arxiv:2409.01927",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-09-03",
      "identifiers": {
        "arxiv": "2409.01927"
      },
      "record_updated_at": "2024-09-03"
    },
    "genre": "preprint",
    "id": "arxiv:2409.01927",
    "issued": {
      "date-parts": [
        [
          2024,
          9,
          3
        ]
      ]
    },
    "note": "First published: 2024-09-03; Recorded arXiv version: v1; Recorded version updated: 2024-09-03T14:17:09Z.",
    "title": "From Grounding to Planning: Benchmarking Bottlenecks in Web Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2410.00079v1",
    "archive": "arXiv",
    "archive_location": "2410.00079",
    "author": [
      {
        "literal": "Wenyue Hua"
      },
      {
        "literal": "Mengting Wan"
      },
      {
        "literal": "Shashank Vadrevu"
      },
      {
        "literal": "Ryan Nadel"
      },
      {
        "literal": "Yongfeng Zhang"
      },
      {
        "literal": "Chi Wang"
      }
    ],
    "citation-key": "arxiv_2410_00079",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "6f4c873d1dd443ffc18fc4b63595fdfc726eb84388eddbfced0dcde67c5e19a7",
        "submitted_at": "2024-09-30T16:52:51Z",
        "updated_at": "2024-09-30T16:52:51Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2410.00079v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2410.00079",
      "catalog_id": "arxiv:2410.00079",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2024-09-30",
      "identifiers": {
        "arxiv": "2410.00079"
      },
      "record_updated_at": "2024-09-30"
    },
    "genre": "preprint",
    "id": "arxiv:2410.00079",
    "issued": {
      "date-parts": [
        [
          2024,
          9,
          30
        ]
      ]
    },
    "note": "First published: 2024-09-30; Recorded arXiv version: v1; Recorded version updated: 2024-09-30T16:52:51Z.",
    "title": "Interactive Speculative Planning: Enhance Agent Efficiency through Co-design of System and User Interface",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2410.00467v3",
    "archive": "arXiv",
    "archive_location": "2410.00467",
    "author": [
      {
        "literal": "Shaoqing Zhang"
      },
      {
        "literal": "Zhuosheng Zhang"
      },
      {
        "literal": "Kehai Chen"
      },
      {
        "literal": "Xinbei Ma"
      },
      {
        "literal": "Muyun Yang"
      },
      {
        "literal": "Tiejun Zhao"
      },
      {
        "literal": "Min Zhang"
      }
    ],
    "citation-key": "arxiv_2410_00467",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "f370d105dc1fc1a377fd1b1fd5d1fc3d0f035694cc7c364e6326c587607c98c6",
        "submitted_at": "2024-10-01T07:49:24Z",
        "updated_at": "2024-12-19T14:29:31Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2410.00467v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2410.00467",
      "catalog_id": "arxiv:2410.00467",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2024-10-01",
      "identifiers": {
        "arxiv": "2410.00467"
      },
      "publication": {
        "authors": [
          "Shaoqing Zhang",
          "Zhuosheng Zhang",
          "Kehai Chen (陈科海)",
          "Xinbei Ma",
          "Muyun Yang (杨沐昀)",
          "Tiejun Zhao (赵铁军)",
          "Min Zhang"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2024/11. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2024,
          11
        ],
        "doi": "10.18653/v1/2024.findings-emnlp.70",
        "event_dates": null,
        "evidence_urls": [
          "https://aclanthology.org/2024.findings-emnlp.70/"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot. Publisher spelling, initials, name annotations or author lists differ from the retained preprint.",
        "pages": "1304-1320",
        "pdf_url": "https://aclanthology.org/2024.findings-emnlp.70.pdf",
        "publisher": "Association for Computational Linguistics",
        "series": null,
        "title": "Dynamic Planning for LLM-based Graphical User Interface Automation",
        "type": "conference-paper",
        "url": "https://aclanthology.org/2024.findings-emnlp.70/",
        "venue": "Findings of the Association for Computational Linguistics: EMNLP 2024",
        "volume": null
      },
      "record_updated_at": "2024-12-19"
    },
    "genre": "preprint",
    "id": "arxiv:2410.00467",
    "issued": {
      "date-parts": [
        [
          2024,
          10,
          1
        ]
      ]
    },
    "note": "First published: 2024-10-01; Recorded arXiv version: v3; Recorded version updated: 2024-12-19T14:29:31Z.",
    "title": "Dynamic Planning for LLM-based Graphical User Interface Automation",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2410.02810v3",
    "archive": "arXiv",
    "archive_location": "2410.02810",
    "author": [
      {
        "literal": "Nikolai Rozanov"
      },
      {
        "literal": "Marek Rei"
      }
    ],
    "citation-key": "arxiv_2410_02810",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "8a2e16e978f50c8d7161a11a11fba6aeda9617663de1746422151d38c4874cc9",
        "submitted_at": "2024-09-21T05:54:35Z",
        "updated_at": "2025-04-08T06:37:51Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2410.02810v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2410.02810",
      "catalog_id": "arxiv:2410.02810",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2024-09-21",
      "identifiers": {
        "arxiv": "2410.02810"
      },
      "record_updated_at": "2025-04-08"
    },
    "genre": "preprint",
    "id": "arxiv:2410.02810",
    "issued": {
      "date-parts": [
        [
          2024,
          9,
          21
        ]
      ]
    },
    "note": "First published: 2024-09-21; Recorded arXiv version: v3; Recorded version updated: 2025-04-08T06:37:51Z.",
    "title": "StateAct: Enhancing LLM Base Agents via Self-prompting and State-tracking",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2410.08164v1",
    "archive": "arXiv",
    "archive_location": "2410.08164",
    "author": [
      {
        "literal": "Saaket Agashe"
      },
      {
        "literal": "Jiuzhou Han"
      },
      {
        "literal": "Shuyu Gan"
      },
      {
        "literal": "Jiachen Yang"
      },
      {
        "literal": "Ang Li"
      },
      {
        "literal": "Xin Eric Wang"
      }
    ],
    "citation-key": "arxiv_2410_08164",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "284888d16b049767f76448f23f5d35fda25b109fd67f450105950173ae81c97b",
        "submitted_at": "2024-10-10T17:43:51Z",
        "updated_at": "2024-10-10T17:43:51Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2410.08164v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2410.08164",
      "catalog_id": "arxiv:2410.08164",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2024-10-10",
      "identifiers": {
        "arxiv": "2410.08164"
      },
      "publication": {
        "authors": [
          "Saaket Agashe",
          "Jiuzhou Han",
          "Shuyu Gan",
          "Jiachen Yang",
          "Ang Li",
          "Xin Wang"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2025-05-01. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2025,
          5,
          1
        ],
        "doi": null,
        "event_dates": null,
        "evidence_urls": [
          "https://proceedings.iclr.cc/paper_files/paper/2025/hash/394c7c30ea87b5c3521b4d9e9d419071-Abstract-Conference.html"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot. Publisher spelling, initials, name annotations or author lists differ from the retained preprint.",
        "pages": "22924-22946",
        "pdf_url": "https://proceedings.iclr.cc/paper_files/paper/2025/file/394c7c30ea87b5c3521b4d9e9d419071-Paper-Conference.pdf",
        "publisher": null,
        "series": null,
        "title": "Agent S: An Open Agentic Framework that Uses Computers Like a Human",
        "type": "conference-paper",
        "url": "https://proceedings.iclr.cc/paper_files/paper/2025/hash/394c7c30ea87b5c3521b4d9e9d419071-Abstract-Conference.html",
        "venue": "International Conference on Learning Representations",
        "volume": "2025"
      },
      "record_updated_at": "2024-10-10"
    },
    "genre": "preprint",
    "id": "arxiv:2410.08164",
    "issued": {
      "date-parts": [
        [
          2024,
          10,
          10
        ]
      ]
    },
    "note": "First published: 2024-10-10; Recorded arXiv version: v1; Recorded version updated: 2024-10-10T17:43:51Z.",
    "title": "Agent S: An Open Agentic Framework that Uses Computers Like a Human",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2410.13232v2",
    "archive": "arXiv",
    "archive_location": "2410.13232",
    "author": [
      {
        "literal": "Hyungjoo Chae"
      },
      {
        "literal": "Namyoung Kim"
      },
      {
        "literal": "Kai Tzu-iunn Ong"
      },
      {
        "literal": "Minju Gwak"
      },
      {
        "literal": "Gwanwoo Song"
      },
      {
        "literal": "Jihoon Kim"
      },
      {
        "literal": "Sunghwan Kim"
      },
      {
        "literal": "Dongha Lee"
      },
      {
        "literal": "Jinyoung Yeo"
      }
    ],
    "citation-key": "arxiv_2410_13232",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "cae86bff4412bbfa7b969ecd60fe7b695f4b7c5ea1be6a67e97919df762d1108",
        "submitted_at": "2024-10-17T05:37:00Z",
        "updated_at": "2025-03-29T08:59:09Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2410.13232v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2410.13232",
      "catalog_id": "arxiv:2410.13232",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-10-17",
      "identifiers": {
        "arxiv": "2410.13232"
      },
      "record_updated_at": "2025-03-29"
    },
    "genre": "preprint",
    "id": "arxiv:2410.13232",
    "issued": {
      "date-parts": [
        [
          2024,
          10,
          17
        ]
      ]
    },
    "note": "First published: 2024-10-17; Recorded arXiv version: v2; Recorded version updated: 2025-03-29T08:59:09Z.",
    "title": "Web Agents with World Models: Learning and Leveraging Environment Dynamics in Web Navigation",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2410.13757v3",
    "archive": "arXiv",
    "archive_location": "2410.13757",
    "author": [
      {
        "literal": "Zichen Zhu"
      },
      {
        "literal": "Hao Tang"
      },
      {
        "literal": "Yansi Li"
      },
      {
        "literal": "Dingye Liu"
      },
      {
        "literal": "Hongshen Xu"
      },
      {
        "literal": "Kunyao Lan"
      },
      {
        "literal": "Danyang Zhang"
      },
      {
        "literal": "Yixuan Jiang"
      },
      {
        "literal": "Hao Zhou"
      },
      {
        "literal": "Chenrun Wang"
      },
      {
        "literal": "Situo Zhang"
      },
      {
        "literal": "Liangtai Sun"
      },
      {
        "literal": "Yixiao Wang"
      },
      {
        "literal": "Yuheng Sun"
      },
      {
        "literal": "Lu Chen"
      },
      {
        "literal": "Kai Yu"
      }
    ],
    "citation-key": "arxiv_2410_13757",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "c0ef02f9c31bf2f095330cf018e7525aecc7644eb1552fb6b511ffced3889122",
        "submitted_at": "2024-10-17T16:53:50Z",
        "updated_at": "2025-05-13T06:25:09Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2410.13757v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2410.13757",
      "catalog_id": "arxiv:2410.13757",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2024-10-17",
      "identifiers": {
        "arxiv": "2410.13757"
      },
      "publication": {
        "authors": [
          "Zichen Zhu",
          "Hao Tang",
          "Yansi Li",
          "Dingye Liu",
          "Hongshen Xu",
          "Kunyao Lan",
          "Danyang Zhang",
          "Yixuan Jiang",
          "Hao Zhou",
          "Chenrun Wang",
          "Situo Zhang",
          "Liangtai Sun",
          "Yixiao Wang",
          "Yuheng Sun",
          "Lu Chen",
          "Kai Yu"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2025/4. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2025,
          4
        ],
        "doi": "10.18653/v1/2025.naacl-demo.43",
        "event_dates": null,
        "evidence_urls": [
          "https://aclanthology.org/2025.naacl-demo.43/"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot.",
        "pages": "535-549",
        "pdf_url": "https://aclanthology.org/2025.naacl-demo.43.pdf",
        "publisher": "Association for Computational Linguistics",
        "series": null,
        "title": "MobA: Multifaceted Memory-Enhanced Adaptive Planning for Efficient Mobile Task Automation",
        "type": "conference-paper",
        "url": "https://aclanthology.org/2025.naacl-demo.43/",
        "venue": "Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (System Demonstrations)",
        "volume": null
      },
      "record_updated_at": "2025-05-13"
    },
    "genre": "preprint",
    "id": "arxiv:2410.13757",
    "issued": {
      "date-parts": [
        [
          2024,
          10,
          17
        ]
      ]
    },
    "note": "First published: 2024-10-17; Recorded arXiv version: v3; Recorded version updated: 2025-05-13T06:25:09Z.",
    "title": "MobA: Multifaceted Memory-Enhanced Adaptive Planning for Efficient Mobile Task Automation",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2410.13825v2",
    "archive": "arXiv",
    "archive_location": "2410.13825",
    "author": [
      {
        "literal": "Ke Yang"
      },
      {
        "literal": "Yao Liu"
      },
      {
        "literal": "Sapana Chaudhary"
      },
      {
        "literal": "Rasool Fakoor"
      },
      {
        "literal": "Pratik Chaudhari"
      },
      {
        "literal": "George Karypis"
      },
      {
        "literal": "Huzefa Rangwala"
      }
    ],
    "citation-key": "arxiv_2410_13825",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "09268d51b3d810ffc0024ae332d840afc4501c8f6a2017128fca525be4809a13",
        "submitted_at": "2024-10-17T17:50:38Z",
        "updated_at": "2025-05-24T03:55:22Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2410.13825v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2410.13825",
      "catalog_id": "arxiv:2410.13825",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-10-17",
      "identifiers": {
        "arxiv": "2410.13825"
      },
      "record_updated_at": "2025-05-24"
    },
    "genre": "preprint",
    "id": "arxiv:2410.13825",
    "issued": {
      "date-parts": [
        [
          2024,
          10,
          17
        ]
      ]
    },
    "note": "First published: 2024-10-17; Recorded arXiv version: v2; Recorded version updated: 2025-05-24T03:55:22Z.",
    "title": "AgentOccam: A Simple Yet Strong Baseline for LLM-Based Web Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2410.18963v1",
    "archive": "arXiv",
    "archive_location": "2410.18963",
    "author": [
      {
        "literal": "Xiaoqiang Wang"
      },
      {
        "literal": "Bang Liu"
      }
    ],
    "citation-key": "arxiv_2410_18963",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "66fa51b108ac79cc43a895ba757ca463d5008bfe6db5cde4bdf3a3bb38c444eb",
        "submitted_at": "2024-10-24T17:58:08Z",
        "updated_at": "2024-10-24T17:58:08Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2410.18963v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2410.18963",
      "catalog_id": "arxiv:2410.18963",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-10-24",
      "identifiers": {
        "arxiv": "2410.18963"
      },
      "record_updated_at": "2024-10-24"
    },
    "genre": "preprint",
    "id": "arxiv:2410.18963",
    "issued": {
      "date-parts": [
        [
          2024,
          10,
          24
        ]
      ]
    },
    "note": "First published: 2024-10-24; Recorded arXiv version: v1; Recorded version updated: 2024-10-24T17:58:08Z.",
    "title": "OSCAR: Operating System Control via State-Aware Reasoning and Re-Planning",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2410.22552v1",
    "archive": "arXiv",
    "archive_location": "2410.22552",
    "author": [
      {
        "literal": "Jaekyeom Kim"
      },
      {
        "literal": "Dong-Ki Kim"
      },
      {
        "literal": "Lajanugen Logeswaran"
      },
      {
        "literal": "Sungryull Sohn"
      },
      {
        "literal": "Honglak Lee"
      }
    ],
    "citation-key": "arxiv_2410_22552",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "bf1fb8aa2e289a72f3c44c99e2d531eaa5eeea08b78ff6f3a21e139d47502b17",
        "submitted_at": "2024-10-29T21:37:04Z",
        "updated_at": "2024-10-29T21:37:04Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2410.22552v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2410.22552",
      "catalog_id": "arxiv:2410.22552",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-10-29",
      "identifiers": {
        "arxiv": "2410.22552"
      },
      "record_updated_at": "2024-10-29"
    },
    "genre": "preprint",
    "id": "arxiv:2410.22552",
    "issued": {
      "date-parts": [
        [
          2024,
          10,
          29
        ]
      ]
    },
    "note": "First published: 2024-10-29; Recorded arXiv version: v1; Recorded version updated: 2024-10-29T21:37:04Z.",
    "title": "Auto-Intent: Automated Intent Discovery and Self-Exploration for Large Language Model Web Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2411.04468v1",
    "archive": "arXiv",
    "archive_location": "2411.04468",
    "author": [
      {
        "literal": "Adam Fourney"
      },
      {
        "literal": "Gagan Bansal"
      },
      {
        "literal": "Hussein Mozannar"
      },
      {
        "literal": "Cheng Tan"
      },
      {
        "literal": "Eduardo Salinas"
      },
      {
        "literal": "Erkang"
      },
      {
        "literal": "Zhu"
      },
      {
        "literal": "Friederike Niedtner"
      },
      {
        "literal": "Grace Proebsting"
      },
      {
        "literal": "Griffin Bassman"
      },
      {
        "literal": "Jack Gerrits"
      },
      {
        "literal": "Jacob Alber"
      },
      {
        "literal": "Peter Chang"
      },
      {
        "literal": "Ricky Loynd"
      },
      {
        "literal": "Robert West"
      },
      {
        "literal": "Victor Dibia"
      },
      {
        "literal": "Ahmed Awadallah"
      },
      {
        "literal": "Ece Kamar"
      },
      {
        "literal": "Rafah Hosn"
      },
      {
        "literal": "Saleema Amershi"
      }
    ],
    "citation-key": "arxiv_2411_04468",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "4a2c499237dbcbe1482824a2710f5185dee05c9ed66a0882d1e422f30afe056a",
        "submitted_at": "2024-11-07T06:36:19Z",
        "updated_at": "2024-11-07T06:36:19Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2411.04468v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2411.04468",
      "catalog_id": "arxiv:2411.04468",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2024-11-07",
      "identifiers": {
        "arxiv": "2411.04468"
      },
      "record_updated_at": "2024-11-07"
    },
    "genre": "preprint",
    "id": "arxiv:2411.04468",
    "issued": {
      "date-parts": [
        [
          2024,
          11,
          7
        ]
      ]
    },
    "note": "First published: 2024-11-07; Recorded arXiv version: v1; Recorded version updated: 2024-11-07T06:36:19Z.",
    "title": "Magentic-One: A Generalist Multi-Agent System for Solving Complex Tasks",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2411.06559v2",
    "archive": "arXiv",
    "archive_location": "2411.06559",
    "author": [
      {
        "literal": "Yu Gu"
      },
      {
        "literal": "Kai Zhang"
      },
      {
        "literal": "Yuting Ning"
      },
      {
        "literal": "Boyuan Zheng"
      },
      {
        "literal": "Boyu Gou"
      },
      {
        "literal": "Tianci Xue"
      },
      {
        "literal": "Cheng Chang"
      },
      {
        "literal": "Sanjari Srivastava"
      },
      {
        "literal": "Yanan Xie"
      },
      {
        "literal": "Peng Qi"
      },
      {
        "literal": "Huan Sun"
      },
      {
        "literal": "Yu Su"
      }
    ],
    "citation-key": "arxiv_2411_06559",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "9ecaca6448aae5c8849d008d820d700bfa438ed0248f85be585b0b831408b9b6",
        "submitted_at": "2024-11-10T18:50:51Z",
        "updated_at": "2025-04-01T05:04:47Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2411.06559v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2411.06559",
      "catalog_id": "arxiv:2411.06559",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-11-10",
      "identifiers": {
        "arxiv": "2411.06559"
      },
      "record_updated_at": "2025-04-01"
    },
    "genre": "preprint",
    "id": "arxiv:2411.06559",
    "issued": {
      "date-parts": [
        [
          2024,
          11,
          10
        ]
      ]
    },
    "note": "First published: 2024-11-10; Recorded arXiv version: v2; Recorded version updated: 2025-04-01T05:04:47Z.",
    "title": "Is Your LLM Secretly a World Model of the Internet? Model-Based Planning for Web Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2411.10323v1",
    "archive": "arXiv",
    "archive_location": "2411.10323",
    "author": [
      {
        "literal": "Siyuan Hu"
      },
      {
        "literal": "Mingyu Ouyang"
      },
      {
        "literal": "Difei Gao"
      },
      {
        "literal": "Mike Zheng Shou"
      }
    ],
    "citation-key": "arxiv_2411_10323",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "255e101e263c4e3a0cb491d4c34b3e75a8f3660cec2141e3a507a30cf5514dbd",
        "submitted_at": "2024-11-15T16:23:52Z",
        "updated_at": "2024-11-15T16:23:52Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2411.10323v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2411.10323",
      "catalog_id": "arxiv:2411.10323",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-11-15",
      "identifiers": {
        "arxiv": "2411.10323"
      },
      "record_updated_at": "2024-11-15"
    },
    "genre": "preprint",
    "id": "arxiv:2411.10323",
    "issued": {
      "date-parts": [
        [
          2024,
          11,
          15
        ]
      ]
    },
    "note": "First published: 2024-11-15; Recorded arXiv version: v1; Recorded version updated: 2024-11-15T16:23:52Z.",
    "title": "The Dawn of GUI Agent: A Preliminary Case Study with Claude 3.5 Computer Use",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2411.10943v1",
    "archive": "arXiv",
    "archive_location": "2411.10943",
    "author": [
      {
        "literal": "Minghe Gao"
      },
      {
        "literal": "Wendong Bu"
      },
      {
        "literal": "Bingchen Miao"
      },
      {
        "literal": "Yang Wu"
      },
      {
        "literal": "Yunfei Li"
      },
      {
        "literal": "Juncheng Li"
      },
      {
        "literal": "Siliang Tang"
      },
      {
        "literal": "Qi Wu"
      },
      {
        "literal": "Yueting Zhuang"
      },
      {
        "literal": "Meng Wang"
      }
    ],
    "citation-key": "arxiv_2411_10943",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "daa9b7092616e94a1d0124d36d8b75d0401b9647ec762c963524616a92def9a5",
        "submitted_at": "2024-11-17T02:44:56Z",
        "updated_at": "2024-11-17T02:44:56Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2411.10943v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2411.10943",
      "catalog_id": "arxiv:2411.10943",
      "catalog_kind": "survey",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-11-17",
      "identifiers": {
        "arxiv": "2411.10943"
      },
      "record_updated_at": "2024-11-17"
    },
    "genre": "preprint",
    "id": "arxiv:2411.10943",
    "issued": {
      "date-parts": [
        [
          2024,
          11,
          17
        ]
      ]
    },
    "note": "First published: 2024-11-17; Recorded arXiv version: v1; Recorded version updated: 2024-11-17T02:44:56Z.",
    "title": "Generalist Virtual Agents: A Survey on Autonomous Agents Across Digital Platforms",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2412.01268v1",
    "archive": "arXiv",
    "archive_location": "2412.01268",
    "author": [
      {
        "literal": "Yiqin Wang"
      },
      {
        "literal": "Haoji Zhang"
      },
      {
        "literal": "Jingqi Tian"
      },
      {
        "literal": "Yansong Tang"
      }
    ],
    "citation-key": "arxiv_2412_01268",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "6a4c750017de421d3ca9bf119834dfb96d2fe8a0610f12b30d2fae603e26c66b",
        "submitted_at": "2024-12-02T08:35:31Z",
        "updated_at": "2024-12-02T08:35:31Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2412.01268v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2412.01268",
      "catalog_id": "arxiv:2412.01268",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2024-12-02",
      "identifiers": {
        "arxiv": "2412.01268"
      },
      "publication": {
        "authors": [
          "Yiqin Wang",
          "Haoji Zhang",
          "Jingqi Tian",
          "Yansong Tang"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2025/7. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2025,
          7
        ],
        "doi": "10.18653/v1/2025.findings-acl.76",
        "event_dates": null,
        "evidence_urls": [
          "https://aclanthology.org/2025.findings-acl.76/"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot.",
        "pages": "1461-1473",
        "pdf_url": "https://aclanthology.org/2025.findings-acl.76.pdf",
        "publisher": "Association for Computational Linguistics",
        "series": null,
        "title": "Ponder & Press: Advancing Visual GUI Agent towards General Computer Control",
        "type": "conference-paper",
        "url": "https://aclanthology.org/2025.findings-acl.76/",
        "venue": "Findings of the Association for Computational Linguistics: ACL 2025",
        "volume": null
      },
      "record_updated_at": "2024-12-02"
    },
    "genre": "preprint",
    "id": "arxiv:2412.01268",
    "issued": {
      "date-parts": [
        [
          2024,
          12,
          2
        ]
      ]
    },
    "note": "First published: 2024-12-02; Recorded arXiv version: v1; Recorded version updated: 2024-12-02T08:35:31Z.",
    "title": "Ponder & Press: Advancing Visual GUI Agent towards General Computer Control",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2412.04454v2",
    "archive": "arXiv",
    "archive_location": "2412.04454",
    "author": [
      {
        "literal": "Yiheng Xu"
      },
      {
        "literal": "Zekun Wang"
      },
      {
        "literal": "Junli Wang"
      },
      {
        "literal": "Dunjie Lu"
      },
      {
        "literal": "Tianbao Xie"
      },
      {
        "literal": "Amrita Saha"
      },
      {
        "literal": "Doyen Sahoo"
      },
      {
        "literal": "Tao Yu"
      },
      {
        "literal": "Caiming Xiong"
      }
    ],
    "citation-key": "arxiv_2412_04454",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "89dbdf874ace87db12fa321e67ecf818d3964622b045ea7d95e109101820d407",
        "submitted_at": "2024-12-05T18:58:26Z",
        "updated_at": "2025-05-05T16:17:20Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2412.04454v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2412.04454",
      "catalog_id": "arxiv:2412.04454",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-12-05",
      "identifiers": {
        "arxiv": "2412.04454"
      },
      "record_updated_at": "2025-05-05"
    },
    "genre": "preprint",
    "id": "arxiv:2412.04454",
    "issued": {
      "date-parts": [
        [
          2024,
          12,
          5
        ]
      ]
    },
    "note": "First published: 2024-12-05; Recorded arXiv version: v2; Recorded version updated: 2025-05-05T16:17:20Z.",
    "title": "Aguvis: Unified Pure Vision Agents for Autonomous GUI Interaction",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2412.17589v1",
    "archive": "arXiv",
    "archive_location": "2412.17589",
    "author": [
      {
        "literal": "Yanheng He"
      },
      {
        "literal": "Jiahe Jin"
      },
      {
        "literal": "Shijie Xia"
      },
      {
        "literal": "Jiadi Su"
      },
      {
        "literal": "Runze Fan"
      },
      {
        "literal": "Haoyang Zou"
      },
      {
        "literal": "Xiangkun Hu"
      },
      {
        "literal": "Pengfei Liu"
      }
    ],
    "citation-key": "arxiv_2412_17589",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "db567d35887fe5367e456b35bc347400b1bc2d147730e9d34c158e1fd7c8dd6c",
        "submitted_at": "2024-12-23T14:02:12Z",
        "updated_at": "2024-12-23T14:02:12Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2412.17589v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2412.17589",
      "catalog_id": "arxiv:2412.17589",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2024-12-23",
      "identifiers": {
        "arxiv": "2412.17589"
      },
      "record_updated_at": "2024-12-23"
    },
    "genre": "preprint",
    "id": "arxiv:2412.17589",
    "issued": {
      "date-parts": [
        [
          2024,
          12,
          23
        ]
      ]
    },
    "note": "First published: 2024-12-23; Recorded arXiv version: v1; Recorded version updated: 2024-12-23T14:02:12Z.",
    "title": "PC Agent: While You Sleep, AI Works -- A Cognitive Journey into Digital World",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2501.04575v1",
    "archive": "arXiv",
    "archive_location": "2501.04575",
    "author": [
      {
        "literal": "Yuhang Liu"
      },
      {
        "literal": "Pengxiang Li"
      },
      {
        "literal": "Zishu Wei"
      },
      {
        "literal": "Congkai Xie"
      },
      {
        "literal": "Xueyu Hu"
      },
      {
        "literal": "Xinchen Xu"
      },
      {
        "literal": "Shengyu Zhang"
      },
      {
        "literal": "Xiaotian Han"
      },
      {
        "literal": "Hongxia Yang"
      },
      {
        "literal": "Fei Wu"
      }
    ],
    "citation-key": "arxiv_2501_04575",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "18d5b2e34f18c24e5d60d5564c9b08eced39c1b17bf5dac8d997640786688fba",
        "submitted_at": "2025-01-08T15:45:21Z",
        "updated_at": "2025-01-08T15:45:21Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2501.04575v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2501.04575",
      "catalog_id": "arxiv:2501.04575",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-01-08",
      "identifiers": {
        "arxiv": "2501.04575"
      },
      "record_updated_at": "2025-01-08"
    },
    "genre": "preprint",
    "id": "arxiv:2501.04575",
    "issued": {
      "date-parts": [
        [
          2025,
          1,
          8
        ]
      ]
    },
    "note": "First published: 2025-01-08; Recorded arXiv version: v1; Recorded version updated: 2025-01-08T15:45:21Z.",
    "title": "InfiGUIAgent: A Multimodal Generalist GUI Agent with Native Reasoning and Reflection",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2501.11425v3",
    "archive": "arXiv",
    "archive_location": "2501.11425",
    "author": [
      {
        "literal": "Siyu Yuan"
      },
      {
        "literal": "Zehui Chen"
      },
      {
        "literal": "Zhiheng Xi"
      },
      {
        "literal": "Junjie Ye"
      },
      {
        "literal": "Zhengyin Du"
      },
      {
        "literal": "Jiecao Chen"
      }
    ],
    "citation-key": "arxiv_2501_11425",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "692a6984b464367c99beb9c5fbd4bbd8378e79f4a4b749503c8fbb72dd80b16f",
        "submitted_at": "2025-01-20T11:46:04Z",
        "updated_at": "2025-03-24T10:18:56Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2501.11425v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2501.11425",
      "catalog_id": "arxiv:2501.11425",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2025-01-20",
      "identifiers": {
        "arxiv": "2501.11425"
      },
      "record_updated_at": "2025-03-24"
    },
    "genre": "preprint",
    "id": "arxiv:2501.11425",
    "issued": {
      "date-parts": [
        [
          2025,
          1,
          20
        ]
      ]
    },
    "note": "First published: 2025-01-20; Recorded arXiv version: v3; Recorded version updated: 2025-03-24T10:18:56Z.",
    "title": "Agent-R: Training Language Model Agents to Reflect via Iterative Self-Training",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2501.12326v1",
    "archive": "arXiv",
    "archive_location": "2501.12326",
    "author": [
      {
        "literal": "Yujia Qin"
      },
      {
        "literal": "Yining Ye"
      },
      {
        "literal": "Junjie Fang"
      },
      {
        "literal": "Haoming Wang"
      },
      {
        "literal": "Shihao Liang"
      },
      {
        "literal": "Shizuo Tian"
      },
      {
        "literal": "Junda Zhang"
      },
      {
        "literal": "Jiahao Li"
      },
      {
        "literal": "Yunxin Li"
      },
      {
        "literal": "Shijue Huang"
      },
      {
        "literal": "Wanjun Zhong"
      },
      {
        "literal": "Kuanye Li"
      },
      {
        "literal": "Jiale Yang"
      },
      {
        "literal": "Yu Miao"
      },
      {
        "literal": "Woyu Lin"
      },
      {
        "literal": "Longxiang Liu"
      },
      {
        "literal": "Xu Jiang"
      },
      {
        "literal": "Qianli Ma"
      },
      {
        "literal": "Jingyu Li"
      },
      {
        "literal": "Xiaojun Xiao"
      },
      {
        "literal": "Kai Cai"
      },
      {
        "literal": "Chuang Li"
      },
      {
        "literal": "Yaowei Zheng"
      },
      {
        "literal": "Chaolin Jin"
      },
      {
        "literal": "Chen Li"
      },
      {
        "literal": "Xiao Zhou"
      },
      {
        "literal": "Minchao Wang"
      },
      {
        "literal": "Haoli Chen"
      },
      {
        "literal": "Zhaojian Li"
      },
      {
        "literal": "Haihua Yang"
      },
      {
        "literal": "Haifeng Liu"
      },
      {
        "literal": "Feng Lin"
      },
      {
        "literal": "Tao Peng"
      },
      {
        "literal": "Xin Liu"
      },
      {
        "literal": "Guang Shi"
      }
    ],
    "citation-key": "arxiv_2501_12326",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d19aa758e66411a31dd1fea6f07c83d1c828bd4aad86f1564a9d7f3670c3f79f",
        "submitted_at": "2025-01-21T17:48:10Z",
        "updated_at": "2025-01-21T17:48:10Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2501.12326v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2501.12326",
      "catalog_id": "arxiv:2501.12326",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-01-21",
      "identifiers": {
        "arxiv": "2501.12326"
      },
      "record_updated_at": "2025-01-21"
    },
    "genre": "preprint",
    "id": "arxiv:2501.12326",
    "issued": {
      "date-parts": [
        [
          2025,
          1,
          21
        ]
      ]
    },
    "note": "First published: 2025-01-21; Recorded arXiv version: v1; Recorded version updated: 2025-01-21T17:48:10Z.",
    "title": "UI-TARS: Pioneering Automated GUI Interaction with Native Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2502.02955v1",
    "archive": "arXiv",
    "archive_location": "2502.02955",
    "author": [
      {
        "literal": "Qinzhuo Wu"
      },
      {
        "literal": "Wei Liu"
      },
      {
        "literal": "Jian Luan"
      },
      {
        "literal": "Bin Wang"
      }
    ],
    "citation-key": "arxiv_2502_02955",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d3e0b19f259bcdb57b0dea92578d0ee1c7c229ea11e622c56e808f17ca2568cc",
        "submitted_at": "2025-02-05T07:35:23Z",
        "updated_at": "2025-02-05T07:35:23Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2502.02955v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2502.02955",
      "catalog_id": "arxiv:2502.02955",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2025-02-05",
      "identifiers": {
        "arxiv": "2502.02955"
      },
      "publication": {
        "authors": [
          "Qinzhuo Wu",
          "Wei Liu",
          "Jian Luan",
          "Bin Wang"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2025/4. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2025,
          4
        ],
        "doi": "10.18653/v1/2025.naacl-long.244",
        "event_dates": null,
        "evidence_urls": [
          "https://aclanthology.org/2025.naacl-long.244/"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot.",
        "pages": "4760-4775",
        "pdf_url": "https://aclanthology.org/2025.naacl-long.244.pdf",
        "publisher": "Association for Computational Linguistics",
        "series": null,
        "title": "ReachAgent: Enhancing Mobile Agent via Page Reaching and Operation",
        "type": "conference-paper",
        "url": "https://aclanthology.org/2025.naacl-long.244/",
        "venue": "Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)",
        "volume": null
      },
      "record_updated_at": "2025-02-05"
    },
    "genre": "preprint",
    "id": "arxiv:2502.02955",
    "issued": {
      "date-parts": [
        [
          2025,
          2,
          5
        ]
      ]
    },
    "note": "First published: 2025-02-05; Recorded arXiv version: v1; Recorded version updated: 2025-02-05T07:35:23Z.",
    "title": "ReachAgent: Enhancing Mobile Agent via Page Reaching and Operation",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2502.07949v2",
    "archive": "arXiv",
    "archive_location": "2502.07949",
    "author": [
      {
        "literal": "Qingyuan Wu"
      },
      {
        "literal": "Jianheng Liu"
      },
      {
        "literal": "Jianye Hao"
      },
      {
        "literal": "Jun Wang"
      },
      {
        "literal": "Kun Shao"
      }
    ],
    "citation-key": "arxiv_2502_07949",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "186b7fd8e79db290ea481645d8adab3989a12f787a13ecb11270f4756f5af42a",
        "submitted_at": "2025-02-11T20:57:46Z",
        "updated_at": "2025-05-20T19:54:36Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2502.07949v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2502.07949",
      "catalog_id": "arxiv:2502.07949",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-02-11",
      "identifiers": {
        "arxiv": "2502.07949"
      },
      "record_updated_at": "2025-05-20"
    },
    "genre": "preprint",
    "id": "arxiv:2502.07949",
    "issued": {
      "date-parts": [
        [
          2025,
          2,
          11
        ]
      ]
    },
    "note": "First published: 2025-02-11; Recorded arXiv version: v2; Recorded version updated: 2025-05-20T19:54:36Z.",
    "title": "Advancing Autonomous VLM Agents via Variational Subgoal-Conditioned Reinforcement Learning",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2502.08047v5",
    "archive": "arXiv",
    "archive_location": "2502.08047",
    "author": [
      {
        "literal": "Henry Hengyuan Zhao"
      },
      {
        "literal": "Kaiming Yang"
      },
      {
        "literal": "Wendi Yu"
      },
      {
        "literal": "Difei Gao"
      },
      {
        "literal": "Mike Zheng Shou"
      }
    ],
    "citation-key": "arxiv_2502_08047",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "68e6ecfcdff0163debe96f6440fb33529356bf9fd2c8743b5953ae52ae71c9ef",
        "submitted_at": "2025-02-12T01:06:10Z",
        "updated_at": "2026-05-23T20:19:43Z",
        "version": 5,
        "versioned_url": "https://arxiv.org/abs/2502.08047v5"
      },
      "canonical_source_url": "https://arxiv.org/abs/2502.08047",
      "catalog_id": "arxiv:2502.08047",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-02-12",
      "identifiers": {
        "arxiv": "2502.08047"
      },
      "record_updated_at": "2026-05-23"
    },
    "genre": "preprint",
    "id": "arxiv:2502.08047",
    "issued": {
      "date-parts": [
        [
          2025,
          2,
          12
        ]
      ]
    },
    "note": "First published: 2025-02-12; Recorded arXiv version: v5; Recorded version updated: 2026-05-23T20:19:43Z.",
    "title": "WorldGUI: An Interactive Benchmark for Desktop GUI Automation from Any Starting Point",
    "type": "article",
    "version": "5"
  },
  {
    "URL": "https://arxiv.org/abs/2502.14282v2",
    "archive": "arXiv",
    "archive_location": "2502.14282",
    "author": [
      {
        "literal": "Haowei Liu"
      },
      {
        "literal": "Xi Zhang"
      },
      {
        "literal": "Haiyang Xu"
      },
      {
        "literal": "Yuyang Wanyan"
      },
      {
        "literal": "Junyang Wang"
      },
      {
        "literal": "Ming Yan"
      },
      {
        "literal": "Ji Zhang"
      },
      {
        "literal": "Chunfeng Yuan"
      },
      {
        "literal": "Changsheng Xu"
      },
      {
        "literal": "Weiming Hu"
      },
      {
        "literal": "Fei Huang"
      }
    ],
    "citation-key": "arxiv_2502_14282",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d75b417f4248397b73934f8b8e45199006e4db6e2ee69b62e0846621411b2dd7",
        "submitted_at": "2025-02-20T05:41:55Z",
        "updated_at": "2025-02-21T02:54:09Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2502.14282v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2502.14282",
      "catalog_id": "arxiv:2502.14282",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-02-20",
      "identifiers": {
        "arxiv": "2502.14282"
      },
      "record_updated_at": "2025-02-21"
    },
    "genre": "preprint",
    "id": "arxiv:2502.14282",
    "issued": {
      "date-parts": [
        [
          2025,
          2,
          20
        ]
      ]
    },
    "note": "First published: 2025-02-20; Recorded arXiv version: v2; Recorded version updated: 2025-02-21T02:54:09Z.",
    "title": "PC-Agent: A Hierarchical Multi-Agent Collaboration Framework for Complex Task Automation on PC",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2503.00401v2",
    "archive": "arXiv",
    "archive_location": "2503.00401",
    "author": [
      {
        "literal": "Zongru Wu"
      },
      {
        "literal": "Pengzhou Cheng"
      },
      {
        "literal": "Zheng Wu"
      },
      {
        "literal": "Tianjie Ju"
      },
      {
        "literal": "Zhuosheng Zhang"
      },
      {
        "literal": "Gongshen Liu"
      }
    ],
    "citation-key": "arxiv_2503_00401",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "53781edbb107437b258d0c0e298f80ec14f8f32b51bc0dd11c0417473eb2c606",
        "submitted_at": "2025-03-01T08:29:59Z",
        "updated_at": "2025-03-04T12:04:26Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2503.00401v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2503.00401",
      "catalog_id": "arxiv:2503.00401",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-03-01",
      "identifiers": {
        "arxiv": "2503.00401"
      },
      "record_updated_at": "2025-03-04"
    },
    "genre": "preprint",
    "id": "arxiv:2503.00401",
    "issued": {
      "date-parts": [
        [
          2025,
          3,
          1
        ]
      ]
    },
    "note": "First published: 2025-03-01; Recorded arXiv version: v2; Recorded version updated: 2025-03-04T12:04:26Z.",
    "title": "Smoothing Grounding and Reasoning for MLLM-Powered GUI Agents with Query-Oriented Pivot Tasks",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2503.02197v2",
    "archive": "arXiv",
    "archive_location": "2503.02197",
    "author": [
      {
        "literal": "Zhixun Chen"
      },
      {
        "literal": "Ming Li"
      },
      {
        "literal": "Yuxuan Huang"
      },
      {
        "literal": "Yali Du"
      },
      {
        "literal": "Meng Fang"
      },
      {
        "literal": "Tianyi Zhou"
      }
    ],
    "citation-key": "arxiv_2503_02197",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "2a58e33cd2b60a8c734968dd92f031b9aa4811932b1a666d947b20ec4c6035a7",
        "submitted_at": "2025-03-04T02:14:55Z",
        "updated_at": "2025-06-05T01:42:08Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2503.02197v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2503.02197",
      "catalog_id": "arxiv:2503.02197",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2025-03-04",
      "identifiers": {
        "arxiv": "2503.02197"
      },
      "record_updated_at": "2025-06-05"
    },
    "genre": "preprint",
    "id": "arxiv:2503.02197",
    "issued": {
      "date-parts": [
        [
          2025,
          3,
          4
        ]
      ]
    },
    "note": "First published: 2025-03-04; Recorded arXiv version: v2; Recorded version updated: 2025-06-05T01:42:08Z.",
    "title": "ATLaS: Agent Tuning via Learning Critical Steps",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2503.02682v2",
    "archive": "arXiv",
    "archive_location": "2503.02682",
    "author": [
      {
        "literal": "Weimin Xiong"
      },
      {
        "literal": "Yifan Song"
      },
      {
        "literal": "Qingxiu Dong"
      },
      {
        "literal": "Bingchan Zhao"
      },
      {
        "literal": "Feifan Song"
      },
      {
        "literal": "Xun Wang"
      },
      {
        "literal": "Sujian Li"
      }
    ],
    "citation-key": "arxiv_2503_02682",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "3395107d8c7db3562bf75a3c22fee55d3da0564bd2f413491bc0d381fc601323",
        "submitted_at": "2025-03-04T14:54:45Z",
        "updated_at": "2025-09-10T16:45:42Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2503.02682v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2503.02682",
      "catalog_id": "arxiv:2503.02682",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2025-03-04",
      "identifiers": {
        "arxiv": "2503.02682"
      },
      "record_updated_at": "2025-09-10"
    },
    "genre": "preprint",
    "id": "arxiv:2503.02682",
    "issued": {
      "date-parts": [
        [
          2025,
          3,
          4
        ]
      ]
    },
    "note": "First published: 2025-03-04; Recorded arXiv version: v2; Recorded version updated: 2025-09-10T16:45:42Z.",
    "title": "MPO: Boosting LLM Agents with Meta Plan Optimization",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2503.02950v2",
    "archive": "arXiv",
    "archive_location": "2503.02950",
    "author": [
      {
        "literal": "Danqing Zhang"
      },
      {
        "literal": "Balaji Rama"
      },
      {
        "literal": "Jingyi Ni"
      },
      {
        "literal": "Shiying He"
      },
      {
        "literal": "Fu Zhao"
      },
      {
        "literal": "Kunyu Chen"
      },
      {
        "literal": "Arnold Chen"
      },
      {
        "literal": "Junyu Cao"
      }
    ],
    "citation-key": "arxiv_2503_02950",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "df682f883bed5e0e9917074dfa130839fabaa318c55714f26943d944774854fe",
        "submitted_at": "2025-03-04T19:13:10Z",
        "updated_at": "2025-05-06T06:42:34Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2503.02950v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2503.02950",
      "catalog_id": "arxiv:2503.02950",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2025-03-04",
      "identifiers": {
        "arxiv": "2503.02950"
      },
      "publication": {
        "authors": [
          "Danqing Zhang",
          "Balaji Rama",
          "Jingyi Ni",
          "Shiying He",
          "Fu Zhao",
          "Kunyu Chen",
          "Arnold Chen",
          "Junyu Cao"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2025/4. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2025,
          4
        ],
        "doi": "10.18653/v1/2025.naacl-demo.36",
        "event_dates": null,
        "evidence_urls": [
          "https://aclanthology.org/2025.naacl-demo.36/"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot.",
        "pages": "449-455",
        "pdf_url": "https://aclanthology.org/2025.naacl-demo.36.pdf",
        "publisher": "Association for Computational Linguistics",
        "series": null,
        "title": "LiteWebAgent: The Open-Source Suite for VLM-Based Web-Agent Applications",
        "type": "conference-paper",
        "url": "https://aclanthology.org/2025.naacl-demo.36/",
        "venue": "Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (System Demonstrations)",
        "volume": null
      },
      "record_updated_at": "2025-05-06"
    },
    "genre": "preprint",
    "id": "arxiv:2503.02950",
    "issued": {
      "date-parts": [
        [
          2025,
          3,
          4
        ]
      ]
    },
    "note": "First published: 2025-03-04; Recorded arXiv version: v2; Recorded version updated: 2025-05-06T06:42:34Z.",
    "title": "LiteWebAgent: The Open-Source Suite for VLM-Based Web-Agent Applications",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2503.03743v1",
    "archive": "arXiv",
    "archive_location": "2503.03743",
    "author": [
      {
        "literal": "Yuqi Zhou"
      },
      {
        "literal": "Shuai Wang"
      },
      {
        "literal": "Sunhao Dai"
      },
      {
        "literal": "Qinglin Jia"
      },
      {
        "literal": "Zhaocheng Du"
      },
      {
        "literal": "Zhenhua Dong"
      },
      {
        "literal": "Jun Xu"
      }
    ],
    "citation-key": "arxiv_2503_03743",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "380d00c46e5dd221b7c481a68b4fbdc00826e64e4b70a98f06a56f8c12b8f5e4",
        "submitted_at": "2025-03-05T18:56:16Z",
        "updated_at": "2025-03-05T18:56:16Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2503.03743v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2503.03743",
      "catalog_id": "arxiv:2503.03743",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-03-05",
      "identifiers": {
        "arxiv": "2503.03743"
      },
      "record_updated_at": "2025-03-05"
    },
    "genre": "preprint",
    "id": "arxiv:2503.03743",
    "issued": {
      "date-parts": [
        [
          2025,
          3,
          5
        ]
      ]
    },
    "note": "First published: 2025-03-05; Recorded arXiv version: v1; Recorded version updated: 2025-03-05T18:56:16Z.",
    "title": "CHOP: Mobile Operating Assistant with Constrained High-frequency Optimized Subtask Planning",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2503.06470v1",
    "archive": "arXiv",
    "archive_location": "2503.06470",
    "author": [
      {
        "literal": "Fei Tang"
      },
      {
        "literal": "Yongliang Shen"
      },
      {
        "literal": "Hang Zhang"
      },
      {
        "literal": "Siqi Chen"
      },
      {
        "literal": "Guiyang Hou"
      },
      {
        "literal": "Wenqi Zhang"
      },
      {
        "literal": "Wenqiao Zhang"
      },
      {
        "literal": "Kaitao Song"
      },
      {
        "literal": "Weiming Lu"
      },
      {
        "literal": "Yueting Zhuang"
      }
    ],
    "citation-key": "arxiv_2503_06470",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "ebc104ee6a868092ae0215f84a3c5c71894bfb321671476a74b794e1645210c7",
        "submitted_at": "2025-03-09T06:14:17Z",
        "updated_at": "2025-03-09T06:14:17Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2503.06470v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2503.06470",
      "catalog_id": "arxiv:2503.06470",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-03-09",
      "identifiers": {
        "arxiv": "2503.06470"
      },
      "record_updated_at": "2025-03-09"
    },
    "genre": "preprint",
    "id": "arxiv:2503.06470",
    "issued": {
      "date-parts": [
        [
          2025,
          3,
          9
        ]
      ]
    },
    "note": "First published: 2025-03-09; Recorded arXiv version: v1; Recorded version updated: 2025-03-09T06:14:17Z.",
    "title": "Think Twice, Click Once: Enhancing GUI Grounding via Fast and Slow Systems",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2503.09263v1",
    "archive": "arXiv",
    "archive_location": "2503.09263",
    "author": [
      {
        "literal": "Di Zhao"
      },
      {
        "literal": "Longhui Ma"
      },
      {
        "literal": "Siwei Wang"
      },
      {
        "literal": "Miao Wang"
      },
      {
        "literal": "Zhao Lv"
      }
    ],
    "citation-key": "arxiv_2503_09263",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "3ea50d595c5d82b4a5d538b4b20725fa1f5afe5b29b03c022b726c307c8cca6e",
        "submitted_at": "2025-03-12T11:03:27Z",
        "updated_at": "2025-03-12T11:03:27Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2503.09263v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2503.09263",
      "catalog_id": "arxiv:2503.09263",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-03-12",
      "identifiers": {
        "arxiv": "2503.09263"
      },
      "record_updated_at": "2025-03-12"
    },
    "genre": "preprint",
    "id": "arxiv:2503.09263",
    "issued": {
      "date-parts": [
        [
          2025,
          3,
          12
        ]
      ]
    },
    "note": "First published: 2025-03-12; Recorded arXiv version: v1; Recorded version updated: 2025-03-12T11:03:27Z.",
    "title": "COLA: A Scalable Multi-Agent Framework For Windows UI Task Automation",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2503.09572v3",
    "archive": "arXiv",
    "archive_location": "2503.09572",
    "author": [
      {
        "literal": "Lutfi Eren Erdogan"
      },
      {
        "literal": "Nicholas Lee"
      },
      {
        "literal": "Sehoon Kim"
      },
      {
        "literal": "Suhong Moon"
      },
      {
        "literal": "Hiroki Furuta"
      },
      {
        "literal": "Gopala Anumanchipalli"
      },
      {
        "literal": "Kurt Keutzer"
      },
      {
        "literal": "Amir Gholami"
      }
    ],
    "citation-key": "arxiv_2503_09572",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "6b03d84f29bb524d7950039624e0d460d8381f2ba63d70c69f5f958013730db9",
        "submitted_at": "2025-03-12T17:40:52Z",
        "updated_at": "2025-04-22T17:56:22Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2503.09572v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2503.09572",
      "catalog_id": "arxiv:2503.09572",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-03-12",
      "identifiers": {
        "arxiv": "2503.09572"
      },
      "record_updated_at": "2025-04-22"
    },
    "genre": "preprint",
    "id": "arxiv:2503.09572",
    "issued": {
      "date-parts": [
        [
          2025,
          3,
          12
        ]
      ]
    },
    "note": "First published: 2025-03-12; Recorded arXiv version: v3; Recorded version updated: 2025-04-22T17:56:22Z.",
    "title": "Plan-and-Act: Improving Planning of Agents for Long-Horizon Tasks",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2503.15937v5",
    "archive": "arXiv",
    "archive_location": "2503.15937",
    "author": [
      {
        "literal": "Gaole Dai"
      },
      {
        "literal": "Shiqi Jiang"
      },
      {
        "literal": "Ting Cao"
      },
      {
        "literal": "Yuanchun Li"
      },
      {
        "literal": "Yuqing Yang"
      },
      {
        "literal": "Rui Tan"
      },
      {
        "literal": "Mo Li"
      },
      {
        "literal": "Lili Qiu"
      }
    ],
    "citation-key": "arxiv_2503_15937",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "6449576cb8ba45569b57844d87f22d01645b3639a222c648d914dc38e3421553",
        "submitted_at": "2025-03-20T08:25:00Z",
        "updated_at": "2026-02-21T01:36:43Z",
        "version": 5,
        "versioned_url": "https://arxiv.org/abs/2503.15937v5"
      },
      "canonical_source_url": "https://arxiv.org/abs/2503.15937",
      "catalog_id": "arxiv:2503.15937",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-03-20",
      "identifiers": {
        "arxiv": "2503.15937"
      },
      "record_updated_at": "2026-02-21"
    },
    "genre": "preprint",
    "id": "arxiv:2503.15937",
    "issued": {
      "date-parts": [
        [
          2025,
          3,
          20
        ]
      ]
    },
    "note": "First published: 2025-03-20; Recorded arXiv version: v5; Recorded version updated: 2026-02-21T01:36:43Z.",
    "title": "Advancing Mobile GUI Agents: A Verifier-Driven Approach to Practical Deployment",
    "type": "article",
    "version": "5"
  },
  {
    "URL": "https://arxiv.org/abs/2503.16788v1",
    "archive": "arXiv",
    "archive_location": "2503.16788",
    "author": [
      {
        "literal": "Li Zhang"
      },
      {
        "literal": "Longxi Gao"
      },
      {
        "literal": "Mengwei Xu"
      }
    ],
    "citation-key": "arxiv_2503_16788",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "e450216f8f802d8f3d16c5886097cef801d051951af0c57fdf195cc176ac744f",
        "submitted_at": "2025-03-21T01:52:43Z",
        "updated_at": "2025-03-21T01:52:43Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2503.16788v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2503.16788",
      "catalog_id": "arxiv:2503.16788",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-03-21",
      "identifiers": {
        "arxiv": "2503.16788"
      },
      "record_updated_at": "2025-03-21"
    },
    "genre": "preprint",
    "id": "arxiv:2503.16788",
    "issued": {
      "date-parts": [
        [
          2025,
          3,
          21
        ]
      ]
    },
    "note": "First published: 2025-03-21; Recorded arXiv version: v1; Recorded version updated: 2025-03-21T01:52:43Z.",
    "title": "Does Chain-of-Thought Reasoning Help Mobile GUI Agent? An Empirical Study",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2503.17709v1",
    "archive": "arXiv",
    "archive_location": "2503.17709",
    "author": [
      {
        "literal": "Yuchen Sun"
      },
      {
        "literal": "Shanhui Zhao"
      },
      {
        "literal": "Tao Yu"
      },
      {
        "literal": "Hao Wen"
      },
      {
        "literal": "Samith Va"
      },
      {
        "literal": "Mengwei Xu"
      },
      {
        "literal": "Yuanchun Li"
      },
      {
        "literal": "Chongyang Zhang"
      }
    ],
    "citation-key": "arxiv_2503_17709",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "80bb507bf3cdccf6f023d0c334899d9d7099c362e9f37c314b15ff9388bae65c",
        "submitted_at": "2025-03-22T09:30:37Z",
        "updated_at": "2025-03-22T09:30:37Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2503.17709v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2503.17709",
      "catalog_id": "arxiv:2503.17709",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-03-22",
      "identifiers": {
        "arxiv": "2503.17709"
      },
      "record_updated_at": "2025-03-22"
    },
    "genre": "preprint",
    "id": "arxiv:2503.17709",
    "issued": {
      "date-parts": [
        [
          2025,
          3,
          22
        ]
      ]
    },
    "note": "First published: 2025-03-22; Recorded arXiv version: v1; Recorded version updated: 2025-03-22T09:30:37Z.",
    "title": "GUI-Xplore: Empowering Generalizable GUI Agents with One Exploration",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2503.24180v2",
    "archive": "arXiv",
    "archive_location": "2503.24180",
    "author": [
      {
        "literal": "Ziming Cheng"
      },
      {
        "literal": "Zhiyuan Huang"
      },
      {
        "literal": "Junting Pan"
      },
      {
        "literal": "Zhaohui Hou"
      },
      {
        "literal": "Mingjie Zhan"
      }
    ],
    "citation-key": "arxiv_2503_24180",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "1bb09b54379d2fbc5b489f24a27f6311b91bb0821eefe5187aa160ffb00639d9",
        "submitted_at": "2025-03-31T14:56:24Z",
        "updated_at": "2025-08-24T08:39:18Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2503.24180v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2503.24180",
      "catalog_id": "arxiv:2503.24180",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-03-31",
      "identifiers": {
        "arxiv": "2503.24180"
      },
      "record_updated_at": "2025-08-24"
    },
    "genre": "preprint",
    "id": "arxiv:2503.24180",
    "issued": {
      "date-parts": [
        [
          2025,
          3,
          31
        ]
      ]
    },
    "note": "First published: 2025-03-31; Recorded arXiv version: v2; Recorded version updated: 2025-08-24T08:39:18Z.",
    "title": "Navi-plus: Managing Ambiguous GUI Navigation Tasks with Follow-up Questions",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2504.00906v1",
    "archive": "arXiv",
    "archive_location": "2504.00906",
    "author": [
      {
        "literal": "Saaket Agashe"
      },
      {
        "literal": "Kyle Wong"
      },
      {
        "literal": "Vincent Tu"
      },
      {
        "literal": "Jiachen Yang"
      },
      {
        "literal": "Ang Li"
      },
      {
        "literal": "Xin Eric Wang"
      }
    ],
    "citation-key": "arxiv_2504_00906",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "11958f69503db514a91af8721870a4cbc1fa19dabf94e3f4ce4118d2fad9862a",
        "submitted_at": "2025-04-01T15:40:27Z",
        "updated_at": "2025-04-01T15:40:27Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2504.00906v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2504.00906",
      "catalog_id": "arxiv:2504.00906",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-04-01",
      "identifiers": {
        "arxiv": "2504.00906"
      },
      "record_updated_at": "2025-04-01"
    },
    "genre": "preprint",
    "id": "arxiv:2504.00906",
    "issued": {
      "date-parts": [
        [
          2025,
          4,
          1
        ]
      ]
    },
    "note": "First published: 2025-04-01; Recorded arXiv version: v1; Recorded version updated: 2025-04-01T15:40:27Z.",
    "title": "Agent S2: A Compositional Generalist-Specialist Framework for Computer Use Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2504.07079v1",
    "archive": "arXiv",
    "archive_location": "2504.07079",
    "author": [
      {
        "literal": "Boyuan Zheng"
      },
      {
        "literal": "Michael Y. Fatemi"
      },
      {
        "literal": "Xiaolong Jin"
      },
      {
        "literal": "Zora Zhiruo Wang"
      },
      {
        "literal": "Apurva Gandhi"
      },
      {
        "literal": "Yueqi Song"
      },
      {
        "literal": "Yu Gu"
      },
      {
        "literal": "Jayanth Srinivasa"
      },
      {
        "literal": "Gaowen Liu"
      },
      {
        "literal": "Graham Neubig"
      },
      {
        "literal": "Yu Su"
      }
    ],
    "citation-key": "arxiv_2504_07079",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "5f0489433a41b49ab8ab42eea57503685ba60beadabea7822134e96d865a7435",
        "submitted_at": "2025-04-09T17:51:50Z",
        "updated_at": "2025-04-09T17:51:50Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2504.07079v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2504.07079",
      "catalog_id": "arxiv:2504.07079",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-04-09",
      "identifiers": {
        "arxiv": "2504.07079"
      },
      "record_updated_at": "2025-04-09"
    },
    "genre": "preprint",
    "id": "arxiv:2504.07079",
    "issued": {
      "date-parts": [
        [
          2025,
          4,
          9
        ]
      ]
    },
    "note": "First published: 2025-04-09; Recorded arXiv version: v1; Recorded version updated: 2025-04-09T17:51:50Z.",
    "title": "SkillWeaver: Web Agents can Self-Improve by Discovering and Honing Skills",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2504.09352v1",
    "archive": "arXiv",
    "archive_location": "2504.09352",
    "author": [
      {
        "literal": "Iason Chaimalas"
      },
      {
        "literal": "Arnas Vyšniauskas"
      },
      {
        "literal": "Gabriel Brostow"
      }
    ],
    "citation-key": "arxiv_2504_09352",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "bf23700bf7626733b0016b773852455c503617adc81025fc9e097879b5534cca",
        "submitted_at": "2025-04-12T22:02:29Z",
        "updated_at": "2025-04-12T22:02:29Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2504.09352v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2504.09352",
      "catalog_id": "arxiv:2504.09352",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-04-12",
      "identifiers": {
        "arxiv": "2504.09352"
      },
      "record_updated_at": "2025-04-12"
    },
    "genre": "preprint",
    "id": "arxiv:2504.09352",
    "issued": {
      "date-parts": [
        [
          2025,
          4,
          12
        ]
      ]
    },
    "note": "First published: 2025-04-12; Recorded arXiv version: v1; Recorded version updated: 2025-04-12T22:02:29Z.",
    "title": "Explorer: Robust Collection of Interactable GUI Elements",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2504.10127v2",
    "archive": "arXiv",
    "archive_location": "2504.10127",
    "author": [
      {
        "literal": "Junlei Zhang"
      },
      {
        "literal": "Zichen Ding"
      },
      {
        "literal": "Chang Ma"
      },
      {
        "literal": "Zijie Chen"
      },
      {
        "literal": "Qiushi Sun"
      },
      {
        "literal": "Zhenzhong Lan"
      },
      {
        "literal": "Junxian He"
      }
    ],
    "citation-key": "arxiv_2504_10127",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "171517999164a3581f71cf363ec6eed357b28ca86908cb97e8498c6958762a5f",
        "submitted_at": "2025-04-14T11:35:02Z",
        "updated_at": "2025-04-15T17:13:46Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2504.10127v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2504.10127",
      "catalog_id": "arxiv:2504.10127",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-04-14",
      "identifiers": {
        "arxiv": "2504.10127"
      },
      "record_updated_at": "2025-04-15"
    },
    "genre": "preprint",
    "id": "arxiv:2504.10127",
    "issued": {
      "date-parts": [
        [
          2025,
          4,
          14
        ]
      ]
    },
    "note": "First published: 2025-04-14; Recorded arXiv version: v2; Recorded version updated: 2025-04-15T17:13:46Z.",
    "title": "Breaking the Data Barrier -- Building GUI Agents Through Task Generalization",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2504.11788v3",
    "archive": "arXiv",
    "archive_location": "2504.11788",
    "author": [
      {
        "literal": "Zhisong Zhang"
      },
      {
        "literal": "Tianqing Fang"
      },
      {
        "literal": "Kaixin Ma"
      },
      {
        "literal": "Wenhao Yu"
      },
      {
        "literal": "Hongming Zhang"
      },
      {
        "literal": "Haitao Mi"
      },
      {
        "literal": "Dong Yu"
      }
    ],
    "citation-key": "arxiv_2504_11788",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "76f84d5375ab7321512a1f4f8724c011cf180bf5a7a0d532ff3218cc9d7a6e75",
        "submitted_at": "2025-04-16T05:41:20Z",
        "updated_at": "2026-01-15T02:00:45Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2504.11788v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2504.11788",
      "catalog_id": "arxiv:2504.11788",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-04-16",
      "identifiers": {
        "arxiv": "2504.11788"
      },
      "record_updated_at": "2026-01-15"
    },
    "genre": "preprint",
    "id": "arxiv:2504.11788",
    "issued": {
      "date-parts": [
        [
          2025,
          4,
          16
        ]
      ]
    },
    "note": "First published: 2025-04-16; Recorded arXiv version: v3; Recorded version updated: 2026-01-15T02:00:45Z.",
    "title": "WebRollback: Enhancing Web Agents with Explicit Rollback Mechanisms",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2504.13936v2",
    "archive": "arXiv",
    "archive_location": "2504.13936",
    "author": [
      {
        "literal": "Dezhao Luo"
      },
      {
        "literal": "Bohan Tang"
      },
      {
        "literal": "Kang Li"
      },
      {
        "literal": "Georgios Papoudakis"
      },
      {
        "literal": "Jifei Song"
      },
      {
        "literal": "Shaogang Gong"
      },
      {
        "literal": "Jianye Hao"
      },
      {
        "literal": "Jun Wang"
      },
      {
        "literal": "Kun Shao"
      }
    ],
    "citation-key": "arxiv_2504_13936",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "15e00237ef710c37c0c9f7768d699f57e9d5dcc190e8de7e20e9ec3b21b82750",
        "submitted_at": "2025-04-15T14:03:10Z",
        "updated_at": "2025-05-20T14:11:37Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2504.13936v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2504.13936",
      "catalog_id": "arxiv:2504.13936",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-04-15",
      "identifiers": {
        "arxiv": "2504.13936"
      },
      "record_updated_at": "2025-05-20"
    },
    "genre": "preprint",
    "id": "arxiv:2504.13936",
    "issued": {
      "date-parts": [
        [
          2025,
          4,
          15
        ]
      ]
    },
    "note": "First published: 2025-04-15; Recorded arXiv version: v2; Recorded version updated: 2025-05-20T14:11:37Z.",
    "title": "ViMo: A Generative Visual GUI World Model for App Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2504.14239v1",
    "archive": "arXiv",
    "archive_location": "2504.14239",
    "author": [
      {
        "literal": "Yuhang Liu"
      },
      {
        "literal": "Pengxiang Li"
      },
      {
        "literal": "Congkai Xie"
      },
      {
        "literal": "Xavier Hu"
      },
      {
        "literal": "Xiaotian Han"
      },
      {
        "literal": "Shengyu Zhang"
      },
      {
        "literal": "Hongxia Yang"
      },
      {
        "literal": "Fei Wu"
      }
    ],
    "citation-key": "arxiv_2504_14239",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "4ce03185cc73c2723f444bd9729c67da9aab7d96b5ef099fb78e47f58790bc18",
        "submitted_at": "2025-04-19T09:25:55Z",
        "updated_at": "2025-04-19T09:25:55Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2504.14239v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2504.14239",
      "catalog_id": "arxiv:2504.14239",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-04-19",
      "identifiers": {
        "arxiv": "2504.14239"
      },
      "record_updated_at": "2025-04-19"
    },
    "genre": "preprint",
    "id": "arxiv:2504.14239",
    "issued": {
      "date-parts": [
        [
          2025,
          4,
          19
        ]
      ]
    },
    "note": "First published: 2025-04-19; Recorded arXiv version: v1; Recorded version updated: 2025-04-19T09:25:55Z.",
    "title": "InfiGUI-R1: Advancing Multimodal GUI Agents from Reactive Actors to Deliberative Reasoners",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2504.21024v2",
    "archive": "arXiv",
    "archive_location": "2504.21024",
    "author": [
      {
        "literal": "Tianqing Fang"
      },
      {
        "literal": "Hongming Zhang"
      },
      {
        "literal": "Zhisong Zhang"
      },
      {
        "literal": "Kaixin Ma"
      },
      {
        "literal": "Wenhao Yu"
      },
      {
        "literal": "Haitao Mi"
      },
      {
        "literal": "Dong Yu"
      }
    ],
    "citation-key": "arxiv_2504_21024",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "86ca58c92ffc2cbbcb0297d8c7618dfab966e36a07b141c068998b0f2934d11e",
        "submitted_at": "2025-04-23T02:54:31Z",
        "updated_at": "2025-08-21T05:55:43Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2504.21024v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2504.21024",
      "catalog_id": "arxiv:2504.21024",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-04-23",
      "identifiers": {
        "arxiv": "2504.21024"
      },
      "record_updated_at": "2025-08-21"
    },
    "genre": "preprint",
    "id": "arxiv:2504.21024",
    "issued": {
      "date-parts": [
        [
          2025,
          4,
          23
        ]
      ]
    },
    "note": "First published: 2025-04-23; Recorded arXiv version: v2; Recorded version updated: 2025-08-21T05:55:43Z.",
    "title": "WebEvolver: Enhancing Web Agent Self-Improvement with Coevolving World Model",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2504.21776v2",
    "archive": "arXiv",
    "archive_location": "2504.21776",
    "author": [
      {
        "literal": "Xiaoxi Li"
      },
      {
        "literal": "Jiajie Jin"
      },
      {
        "literal": "Guanting Dong"
      },
      {
        "literal": "Hongjin Qian"
      },
      {
        "literal": "Yongkang Wu"
      },
      {
        "literal": "Ji-Rong Wen"
      },
      {
        "literal": "Yutao Zhu"
      },
      {
        "literal": "Zhicheng Dou"
      }
    ],
    "citation-key": "arxiv_2504_21776",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "351b6d7fbe4e217a5211884832c6ad07d76a0463e6a8c996815999ff3a5f1921",
        "submitted_at": "2025-04-30T16:25:25Z",
        "updated_at": "2025-10-13T12:40:15Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2504.21776v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2504.21776",
      "catalog_id": "arxiv:2504.21776",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2025-04-30",
      "identifiers": {
        "arxiv": "2504.21776"
      },
      "record_updated_at": "2025-10-13"
    },
    "genre": "preprint",
    "id": "arxiv:2504.21776",
    "issued": {
      "date-parts": [
        [
          2025,
          4,
          30
        ]
      ]
    },
    "note": "First published: 2025-04-30; Recorded arXiv version: v2; Recorded version updated: 2025-10-13T12:40:15Z.",
    "title": "WebThinker: Empowering Large Reasoning Models with Deep Research Capability",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2505.00416v1",
    "archive": "arXiv",
    "archive_location": "2505.00416",
    "author": [
      {
        "literal": "Jing Huang"
      },
      {
        "literal": "Zhixiong Zeng"
      },
      {
        "literal": "Wenkang Han"
      },
      {
        "literal": "Yufeng Zhong"
      },
      {
        "literal": "Liming Zheng"
      },
      {
        "literal": "Shuai Fu"
      },
      {
        "literal": "Jingyuan Chen"
      },
      {
        "literal": "Lin Ma"
      }
    ],
    "citation-key": "arxiv_2505_00416",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "e582c92521ec41a0a3daa0ca72b8c00250621289352b28400ee6d311c4a106eb",
        "submitted_at": "2025-05-01T09:27:13Z",
        "updated_at": "2025-05-01T09:27:13Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2505.00416v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2505.00416",
      "catalog_id": "arxiv:2505.00416",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-05-01",
      "identifiers": {
        "arxiv": "2505.00416"
      },
      "record_updated_at": "2025-05-01"
    },
    "genre": "preprint",
    "id": "arxiv:2505.00416",
    "issued": {
      "date-parts": [
        [
          2025,
          5,
          1
        ]
      ]
    },
    "note": "First published: 2025-05-01; Recorded arXiv version: v1; Recorded version updated: 2025-05-01T09:27:13Z.",
    "title": "ScaleTrack: Scaling and back-tracking Automated GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2505.05440v3",
    "archive": "arXiv",
    "archive_location": "2505.05440",
    "author": [
      {
        "literal": "Biao Yi"
      },
      {
        "literal": "Xavier Hu"
      },
      {
        "literal": "Yurun Chen"
      },
      {
        "literal": "Shengyu Zhang"
      },
      {
        "literal": "Hongxia Yang"
      },
      {
        "literal": "Fan Wu"
      }
    ],
    "citation-key": "arxiv_2505_05440",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "8f5fdf474c1cb0d7e9f42d617de27afae38a9fd8f5e0b07d1c86e082dc872a35",
        "submitted_at": "2025-05-08T17:31:20Z",
        "updated_at": "2025-11-17T09:12:00Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2505.05440v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2505.05440",
      "catalog_id": "arxiv:2505.05440",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-05-08",
      "identifiers": {
        "arxiv": "2505.05440"
      },
      "record_updated_at": "2025-11-17"
    },
    "genre": "preprint",
    "id": "arxiv:2505.05440",
    "issued": {
      "date-parts": [
        [
          2025,
          5,
          8
        ]
      ]
    },
    "note": "First published: 2025-05-08; Recorded arXiv version: v3; Recorded version updated: 2025-11-17T09:12:00Z.",
    "title": "EcoAgent: An Efficient Device-Cloud Collaborative Multi-Agent Framework for Mobile Automation",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2505.12299v4",
    "archive": "arXiv",
    "archive_location": "2505.12299",
    "author": [
      {
        "literal": "Kun Huang"
      },
      {
        "literal": "Weikai Xu"
      },
      {
        "literal": "Yuxuan Liu"
      },
      {
        "literal": "Quandong Wang"
      },
      {
        "literal": "Pengzhi Gao"
      },
      {
        "literal": "Wei Liu"
      },
      {
        "literal": "Jian Luan"
      },
      {
        "literal": "Bin Wang"
      },
      {
        "literal": "Bo An"
      }
    ],
    "citation-key": "arxiv_2505_12299",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "9bc92029e16b5afb4b54837d2ab2304c7b8176548061c9deafdca101754128dc",
        "submitted_at": "2025-05-18T08:28:05Z",
        "updated_at": "2026-03-21T09:25:27Z",
        "version": 4,
        "versioned_url": "https://arxiv.org/abs/2505.12299v4"
      },
      "canonical_source_url": "https://arxiv.org/abs/2505.12299",
      "catalog_id": "arxiv:2505.12299",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-05-18",
      "identifiers": {
        "arxiv": "2505.12299"
      },
      "record_updated_at": "2026-03-21"
    },
    "genre": "preprint",
    "id": "arxiv:2505.12299",
    "issued": {
      "date-parts": [
        [
          2025,
          5,
          18
        ]
      ]
    },
    "note": "First published: 2025-05-18; Recorded arXiv version: v4; Recorded version updated: 2026-03-21T09:25:27Z.",
    "title": "MobileIPL: Enhancing Mobile Agents Thinking Process via Iterative Preference Learning",
    "type": "article",
    "version": "4"
  },
  {
    "URL": "https://arxiv.org/abs/2505.14141v1",
    "archive": "arXiv",
    "archive_location": "2505.14141",
    "author": [
      {
        "literal": "Fanglin Mo"
      },
      {
        "literal": "Junzhe Chen"
      },
      {
        "literal": "Haoxuan Zhu"
      },
      {
        "literal": "Xuming Hu"
      }
    ],
    "citation-key": "arxiv_2505_14141",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "de6c501a42c3828bdf795ac59c38b982e738c38bd5d0c59cfb6cb10b51e39292",
        "submitted_at": "2025-05-20T09:45:55Z",
        "updated_at": "2025-05-20T09:45:55Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2505.14141v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2505.14141",
      "catalog_id": "arxiv:2505.14141",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-05-20",
      "identifiers": {
        "arxiv": "2505.14141"
      },
      "record_updated_at": "2025-05-20"
    },
    "genre": "preprint",
    "id": "arxiv:2505.14141",
    "issued": {
      "date-parts": [
        [
          2025,
          5,
          20
        ]
      ]
    },
    "note": "First published: 2025-05-20; Recorded arXiv version: v1; Recorded version updated: 2025-05-20T09:45:55Z.",
    "title": "Building a Stable Planner: An Extended Finite State Machine Based Planning Module for Mobile GUI Agent",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2505.16422v3",
    "archive": "arXiv",
    "archive_location": "2505.16422",
    "author": [
      {
        "literal": "Xiaoran Yin"
      },
      {
        "literal": "Xu Luo"
      },
      {
        "literal": "Hao Wu"
      },
      {
        "literal": "Lianli Gao"
      },
      {
        "literal": "Jingkuan Song"
      }
    ],
    "citation-key": "arxiv_2505_16422",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "93c67222fb0f581e9f9cb3ff59c8d28faec9dc037180060ad5ac962b4469144c",
        "submitted_at": "2025-05-22T09:08:47Z",
        "updated_at": "2025-09-05T02:24:00Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2505.16422v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2505.16422",
      "catalog_id": "arxiv:2505.16422",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-05-22",
      "identifiers": {
        "arxiv": "2505.16422"
      },
      "record_updated_at": "2025-09-05"
    },
    "genre": "preprint",
    "id": "arxiv:2505.16422",
    "issued": {
      "date-parts": [
        [
          2025,
          5,
          22
        ]
      ]
    },
    "note": "First published: 2025-05-22; Recorded arXiv version: v3; Recorded version updated: 2025-09-05T02:24:00Z.",
    "title": "Unlocking Smarter Device Control: Foresighted Planning with a World Model-Driven Code Execution Approach",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2505.16827v1",
    "archive": "arXiv",
    "archive_location": "2505.16827",
    "author": [
      {
        "literal": "Bin Xie"
      },
      {
        "literal": "Rui Shao"
      },
      {
        "literal": "Gongwei Chen"
      },
      {
        "literal": "Kaiwen Zhou"
      },
      {
        "literal": "Yinchuan Li"
      },
      {
        "literal": "Jie Liu"
      },
      {
        "literal": "Min Zhang"
      },
      {
        "literal": "Liqiang Nie"
      }
    ],
    "citation-key": "arxiv_2505_16827",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "641001c49fccac1e1c0659fe765eba5a30cefd2a7ff5d00c821f79092c45a1df",
        "submitted_at": "2025-05-22T16:01:06Z",
        "updated_at": "2025-05-22T16:01:06Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2505.16827v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2505.16827",
      "catalog_id": "arxiv:2505.16827",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "record",
      "first_published_at": "2025-05-22",
      "identifiers": {
        "arxiv": "2505.16827"
      },
      "publication": {
        "authors": [
          "Bin Xie",
          "Rui Shao",
          "Gongwei Chen",
          "Kaiwen Zhou",
          "Yinchuan Li",
          "Jie Liu",
          "Min Zhang",
          "Liqiang Nie"
        ],
        "date_basis": "Explicit publisher citation_publication_date: 2025/7. Preserve supplied precision; do not substitute conference event dates.",
        "date_parts": [
          2025,
          7
        ],
        "doi": "10.18653/v1/2025.acl-long.282",
        "event_dates": null,
        "evidence_urls": [
          "https://aclanthology.org/2025.acl-long.282/"
        ],
        "identity_basis": "Same work: normalized title, ordered author correspondence and the specific contribution in the complete publisher abstract were reviewed against the retained arXiv identity.",
        "issue": null,
        "notes": "Publisher proceedings edition; retained separately from the arXiv snapshot.",
        "pages": "5650-5667",
        "pdf_url": "https://aclanthology.org/2025.acl-long.282.pdf",
        "publisher": "Association for Computational Linguistics",
        "series": null,
        "title": "GUI-explorer: Autonomous Exploration and Mining of Transition-aware Knowledge for GUI Agent",
        "type": "conference-paper",
        "url": "https://aclanthology.org/2025.acl-long.282/",
        "venue": "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
        "volume": null
      },
      "record_updated_at": "2025-05-22"
    },
    "genre": "preprint",
    "id": "arxiv:2505.16827",
    "issued": {
      "date-parts": [
        [
          2025,
          5,
          22
        ]
      ]
    },
    "note": "First published: 2025-05-22; Recorded arXiv version: v1; Recorded version updated: 2025-05-22T16:01:06Z.",
    "title": "GUI-explorer: Autonomous Exploration and Mining of Transition-aware Knowledge for GUI Agent",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2505.20013v2",
    "archive": "arXiv",
    "archive_location": "2505.20013",
    "author": [
      {
        "literal": "Minda Hu"
      },
      {
        "literal": "Tianqing Fang"
      },
      {
        "literal": "Jianshu Zhang"
      },
      {
        "literal": "Junyu Ma"
      },
      {
        "literal": "Zhisong Zhang"
      },
      {
        "literal": "Jingyan Zhou"
      },
      {
        "literal": "Hongming Zhang"
      },
      {
        "literal": "Haitao Mi"
      },
      {
        "literal": "Dong Yu"
      },
      {
        "literal": "Irwin King"
      }
    ],
    "citation-key": "arxiv_2505_20013",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "171071f26a6dc8b9c8851d2f11d60b113983191c905631e431f352711e0d0e26",
        "submitted_at": "2025-05-26T14:03:37Z",
        "updated_at": "2025-09-18T11:32:15Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2505.20013v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2505.20013",
      "catalog_id": "arxiv:2505.20013",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-05-26",
      "identifiers": {
        "arxiv": "2505.20013"
      },
      "record_updated_at": "2025-09-18"
    },
    "genre": "preprint",
    "id": "arxiv:2505.20013",
    "issued": {
      "date-parts": [
        [
          2025,
          5,
          26
        ]
      ]
    },
    "note": "First published: 2025-05-26; Recorded arXiv version: v2; Recorded version updated: 2025-09-18T11:32:15Z.",
    "title": "WebCoT: Enhancing Web Agent Reasoning by Reconstructing Chain-of-Thought in Reflection, Branching, and Rollback",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2505.20660v1",
    "archive": "arXiv",
    "archive_location": "2505.20660",
    "author": [
      {
        "literal": "Qinzhuo Wu"
      },
      {
        "literal": "Pengzhi Gao"
      },
      {
        "literal": "Wei Liu"
      },
      {
        "literal": "Jian Luan"
      }
    ],
    "citation-key": "arxiv_2505_20660",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d2d51388193e19501b580740969a6e08661c6e0a2cf09e2489c46fe4f19209cd",
        "submitted_at": "2025-05-27T03:09:06Z",
        "updated_at": "2025-05-27T03:09:06Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2505.20660v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2505.20660",
      "catalog_id": "arxiv:2505.20660",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-05-27",
      "identifiers": {
        "arxiv": "2505.20660"
      },
      "record_updated_at": "2025-05-27"
    },
    "genre": "preprint",
    "id": "arxiv:2505.20660",
    "issued": {
      "date-parts": [
        [
          2025,
          5,
          27
        ]
      ]
    },
    "note": "First published: 2025-05-27; Recorded arXiv version: v1; Recorded version updated: 2025-05-27T03:09:06Z.",
    "title": "BacktrackAgent: Enhancing GUI Agent with Error Detection and Backtracking Mechanism",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2505.23596v3",
    "archive": "arXiv",
    "archive_location": "2505.23596",
    "author": [
      {
        "literal": "Linqiang Guo"
      },
      {
        "literal": "Wei Liu"
      },
      {
        "literal": "Yi Wen Heng"
      },
      {
        "literal": "Tse-Hsun"
      },
      {
        "literal": "Chen"
      },
      {
        "literal": "Yang Wang"
      }
    ],
    "citation-key": "arxiv_2505_23596",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "6f0ec81429fb39e56414d3bfe621df6c6adf461e63fc1a1c8886ea2b6850b8ee",
        "submitted_at": "2025-05-29T16:08:51Z",
        "updated_at": "2025-11-19T16:48:16Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2505.23596v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2505.23596",
      "catalog_id": "arxiv:2505.23596",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-05-29",
      "identifiers": {
        "arxiv": "2505.23596"
      },
      "record_updated_at": "2025-11-19"
    },
    "genre": "preprint",
    "id": "arxiv:2505.23596",
    "issued": {
      "date-parts": [
        [
          2025,
          5,
          29
        ]
      ]
    },
    "note": "First published: 2025-05-29; Recorded arXiv version: v3; Recorded version updated: 2025-11-19T16:48:16Z.",
    "title": "Agent-SAMA: State-Aware Mobile Assistant",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2506.04614v2",
    "archive": "arXiv",
    "archive_location": "2506.04614",
    "author": [
      {
        "literal": "Yuyang Wanyan"
      },
      {
        "literal": "Xi Zhang"
      },
      {
        "literal": "Haiyang Xu"
      },
      {
        "literal": "Haowei Liu"
      },
      {
        "literal": "Junyang Wang"
      },
      {
        "literal": "Jiabo Ye"
      },
      {
        "literal": "Yutong Kou"
      },
      {
        "literal": "Ming Yan"
      },
      {
        "literal": "Fei Huang"
      },
      {
        "literal": "Xiaoshan Yang"
      },
      {
        "literal": "Weiming Dong"
      },
      {
        "literal": "Changsheng Xu"
      }
    ],
    "citation-key": "arxiv_2506_04614",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "42be944224a9b7d2c8e7042968ab31d373a0bafa304c0e669728cb0f10cc813d",
        "submitted_at": "2025-06-05T04:12:36Z",
        "updated_at": "2025-11-17T07:48:59Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2506.04614v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2506.04614",
      "catalog_id": "arxiv:2506.04614",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-06-05",
      "identifiers": {
        "arxiv": "2506.04614"
      },
      "record_updated_at": "2025-11-17"
    },
    "genre": "preprint",
    "id": "arxiv:2506.04614",
    "issued": {
      "date-parts": [
        [
          2025,
          6,
          5
        ]
      ]
    },
    "note": "First published: 2025-06-05; Recorded arXiv version: v2; Recorded version updated: 2025-11-17T07:48:59Z.",
    "title": "Look Before You Leap: A GUI-Critic-R1 Model for Pre-Operative Error Diagnosis in GUI Automation",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2506.05079v4",
    "archive": "arXiv",
    "archive_location": "2506.05079",
    "author": [
      {
        "literal": "Shengcheng Yu"
      },
      {
        "literal": "Yuchen Ling"
      },
      {
        "literal": "Chunrong Fang"
      },
      {
        "literal": "Quan Zhou"
      },
      {
        "literal": "Yi Zhao"
      },
      {
        "literal": "Chunyang Chen"
      },
      {
        "literal": "Shaomin Zhu"
      },
      {
        "literal": "Zhenyu Chen"
      }
    ],
    "citation-key": "arxiv_2506_05079",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "bad447e8bca06f720a9476492de06978c73a356a38b00671203f8af0b81f1952",
        "submitted_at": "2025-06-05T14:27:40Z",
        "updated_at": "2026-05-03T08:25:53Z",
        "version": 4,
        "versioned_url": "https://arxiv.org/abs/2506.05079v4"
      },
      "canonical_source_url": "https://arxiv.org/abs/2506.05079",
      "catalog_id": "arxiv:2506.05079",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-06-05",
      "identifiers": {
        "arxiv": "2506.05079"
      },
      "record_updated_at": "2026-05-03"
    },
    "genre": "preprint",
    "id": "arxiv:2506.05079",
    "issued": {
      "date-parts": [
        [
          2025,
          6,
          5
        ]
      ]
    },
    "note": "First published: 2025-06-05; Recorded arXiv version: v4; Recorded version updated: 2026-05-03T08:25:53Z.",
    "title": "Scenario-Guided LLM-based Mobile App GUI Testing",
    "type": "article",
    "version": "4"
  },
  {
    "URL": "https://arxiv.org/abs/2506.07217v2",
    "archive": "arXiv",
    "archive_location": "2506.07217",
    "author": [
      {
        "literal": "Zihan Deng"
      },
      {
        "literal": "Changyu Du"
      },
      {
        "literal": "Stavros Nousias"
      },
      {
        "literal": "André Borrmann"
      }
    ],
    "citation-key": "arxiv_2506_07217",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "2bf4f91de32fdf3bd9b81ba64ace02e4a7d02ff094794f30dfd560389f334cea",
        "submitted_at": "2025-06-08T16:45:31Z",
        "updated_at": "2025-06-30T08:31:07Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2506.07217v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2506.07217",
      "catalog_id": "arxiv:2506.07217",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-06-08",
      "identifiers": {
        "arxiv": "2506.07217"
      },
      "record_updated_at": "2025-06-30"
    },
    "genre": "preprint",
    "id": "arxiv:2506.07217",
    "issued": {
      "date-parts": [
        [
          2025,
          6,
          8
        ]
      ]
    },
    "note": "First published: 2025-06-08; Recorded arXiv version: v2; Recorded version updated: 2025-06-30T08:31:07Z.",
    "title": "BIMgent: Towards Autonomous Building Modeling via Computer-use Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2506.07976v2",
    "archive": "arXiv",
    "archive_location": "2506.07976",
    "author": [
      {
        "literal": "Junhong Shen"
      },
      {
        "literal": "Hao Bai"
      },
      {
        "literal": "Lunjun Zhang"
      },
      {
        "literal": "Yifei Zhou"
      },
      {
        "literal": "Amrith Setlur"
      },
      {
        "literal": "Shengbang Tong"
      },
      {
        "literal": "Diego Caples"
      },
      {
        "literal": "Nan Jiang"
      },
      {
        "literal": "Tong Zhang"
      },
      {
        "literal": "Ameet Talwalkar"
      },
      {
        "literal": "Aviral Kumar"
      }
    ],
    "citation-key": "arxiv_2506_07976",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "e4b6c8b297e88dae689ba2a48724a441698cd5309a3bbe1198f38343e88aceb1",
        "submitted_at": "2025-06-09T17:50:02Z",
        "updated_at": "2025-06-10T12:50:18Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2506.07976v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2506.07976",
      "catalog_id": "arxiv:2506.07976",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-06-09",
      "identifiers": {
        "arxiv": "2506.07976"
      },
      "record_updated_at": "2025-06-10"
    },
    "genre": "preprint",
    "id": "arxiv:2506.07976",
    "issued": {
      "date-parts": [
        [
          2025,
          6,
          9
        ]
      ]
    },
    "note": "First published: 2025-06-09; Recorded arXiv version: v2; Recorded version updated: 2025-06-10T12:50:18Z.",
    "title": "Thinking vs. Doing: Agents that Reason by Scaling Test-Time Interaction",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2506.08136v3",
    "archive": "arXiv",
    "archive_location": "2506.08136",
    "author": [
      {
        "literal": "Zefang Liu"
      },
      {
        "literal": "Yinzhu Quan"
      }
    ],
    "citation-key": "arxiv_2506_08136",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "43aca3b0b8a1a0d5729a4e90dc1516bbb68e3611de3220876cdb29afc81d847f",
        "submitted_at": "2025-06-09T18:39:48Z",
        "updated_at": "2026-05-11T08:25:57Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2506.08136v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2506.08136",
      "catalog_id": "arxiv:2506.08136",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-06-09",
      "identifiers": {
        "arxiv": "2506.08136"
      },
      "record_updated_at": "2026-05-11"
    },
    "genre": "preprint",
    "id": "arxiv:2506.08136",
    "issued": {
      "date-parts": [
        [
          2025,
          6,
          9
        ]
      ]
    },
    "note": "First published: 2025-06-09; Recorded arXiv version: v3; Recorded version updated: 2026-05-11T08:25:57Z.",
    "title": "EconWebArena: Benchmarking Autonomous Agents on Economic Tasks in Realistic Web Environments",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2506.08972v1",
    "archive": "arXiv",
    "archive_location": "2506.08972",
    "author": [
      {
        "literal": "Yuan Guo"
      },
      {
        "literal": "Tingjia Miao"
      },
      {
        "literal": "Zheng Wu"
      },
      {
        "literal": "Pengzhou Cheng"
      },
      {
        "literal": "Ming Zhou"
      },
      {
        "literal": "Zhuosheng Zhang"
      }
    ],
    "citation-key": "arxiv_2506_08972",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "8dbc84aa582b94775d490cf84e17eff05be84af7dbcc2de7de61b886fa3c9178",
        "submitted_at": "2025-06-10T16:45:29Z",
        "updated_at": "2025-06-10T16:45:29Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2506.08972v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2506.08972",
      "catalog_id": "arxiv:2506.08972",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-06-10",
      "identifiers": {
        "arxiv": "2506.08972"
      },
      "record_updated_at": "2025-06-10"
    },
    "genre": "preprint",
    "id": "arxiv:2506.08972",
    "issued": {
      "date-parts": [
        [
          2025,
          6,
          10
        ]
      ]
    },
    "note": "First published: 2025-06-10; Recorded arXiv version: v1; Recorded version updated: 2025-06-10T16:45:29Z.",
    "title": "Atomic-to-Compositional Generalization for Mobile Agents with A New Benchmark and Scheduling System",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2506.10387v1",
    "archive": "arXiv",
    "archive_location": "2506.10387",
    "author": [
      {
        "literal": "Yuquan Xie"
      },
      {
        "literal": "Zaijing Li"
      },
      {
        "literal": "Rui Shao"
      },
      {
        "literal": "Gongwei Chen"
      },
      {
        "literal": "Kaiwen Zhou"
      },
      {
        "literal": "Yinchuan Li"
      },
      {
        "literal": "Dongmei Jiang"
      },
      {
        "literal": "Liqiang Nie"
      }
    ],
    "citation-key": "arxiv_2506_10387",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "9b353606ce29a5bc60d263e86827a825af46883e25da5bd290b48affa9dd3838",
        "submitted_at": "2025-06-12T06:21:19Z",
        "updated_at": "2025-06-12T06:21:19Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2506.10387v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2506.10387",
      "catalog_id": "arxiv:2506.10387",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-06-12",
      "identifiers": {
        "arxiv": "2506.10387"
      },
      "record_updated_at": "2025-06-12"
    },
    "genre": "preprint",
    "id": "arxiv:2506.10387",
    "issued": {
      "date-parts": [
        [
          2025,
          6,
          12
        ]
      ]
    },
    "note": "First published: 2025-06-12; Recorded arXiv version: v1; Recorded version updated: 2025-06-12T06:21:19Z.",
    "title": "Mirage-1: Augmenting and Updating GUI Agent with Hierarchical Multimodal Skills",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2506.15677v3",
    "archive": "arXiv",
    "archive_location": "2506.15677",
    "author": [
      {
        "literal": "Yining Hong"
      },
      {
        "literal": "Rui Sun"
      },
      {
        "literal": "Bingxuan Li"
      },
      {
        "literal": "Xingcheng Yao"
      },
      {
        "literal": "Maxine Wu"
      },
      {
        "literal": "Alexander Chien"
      },
      {
        "literal": "Da Yin"
      },
      {
        "literal": "Ying Nian Wu"
      },
      {
        "literal": "Zhecan James Wang"
      },
      {
        "literal": "Kai-Wei Chang"
      }
    ],
    "citation-key": "arxiv_2506_15677",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "3e8f714b234a1aad60d0306df23c90410aebfd257fb5906644d0382479e8f3d0",
        "submitted_at": "2025-06-18T17:58:17Z",
        "updated_at": "2025-07-29T22:40:49Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2506.15677v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2506.15677",
      "catalog_id": "arxiv:2506.15677",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-06-18",
      "identifiers": {
        "arxiv": "2506.15677"
      },
      "record_updated_at": "2025-07-29"
    },
    "genre": "preprint",
    "id": "arxiv:2506.15677",
    "issued": {
      "date-parts": [
        [
          2025,
          6,
          18
        ]
      ]
    },
    "note": "First published: 2025-06-18; Recorded arXiv version: v3; Recorded version updated: 2025-07-29T22:40:49Z.",
    "title": "Embodied Web Agents: Bridging Physical-Digital Realms for Integrated Agent Intelligence",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2506.17913v2",
    "archive": "arXiv",
    "archive_location": "2506.17913",
    "author": [
      {
        "literal": "Jinjie Wei"
      },
      {
        "literal": "Jiyao Liu"
      },
      {
        "literal": "Lihao Liu"
      },
      {
        "literal": "Ming Hu"
      },
      {
        "literal": "Junzhi Ning"
      },
      {
        "literal": "Mingcheng Li"
      },
      {
        "literal": "Weijie Yin"
      },
      {
        "literal": "Junjun He"
      },
      {
        "literal": "Xiao Liang"
      },
      {
        "literal": "Chao Feng"
      },
      {
        "literal": "Dingkang Yang"
      }
    ],
    "citation-key": "arxiv_2506_17913",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "f12d2ae58243b5f89a6a510e8f3ef4073a605c4bafdf63c37fd38db57ea84602",
        "submitted_at": "2025-06-22T06:30:52Z",
        "updated_at": "2026-07-21T05:30:48Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2506.17913v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2506.17913",
      "catalog_id": "arxiv:2506.17913",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-06-22",
      "identifiers": {
        "arxiv": "2506.17913"
      },
      "record_updated_at": "2026-07-21"
    },
    "genre": "preprint",
    "id": "arxiv:2506.17913",
    "issued": {
      "date-parts": [
        [
          2025,
          6,
          22
        ]
      ]
    },
    "note": "First published: 2025-06-22; Recorded arXiv version: v2; Recorded version updated: 2026-07-21T05:30:48Z.",
    "title": "Learning, Reasoning, Refinement: A Framework for Kahneman's Dual-System Intelligence in GUI Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2507.00210v1",
    "archive": "arXiv",
    "archive_location": "2507.00210",
    "author": [
      {
        "literal": "Imene Kerboua"
      },
      {
        "literal": "Sahar Omidi Shayegan"
      },
      {
        "literal": "Megh Thakkar"
      },
      {
        "literal": "Xing Han Lù"
      },
      {
        "literal": "Massimo Caccia"
      },
      {
        "literal": "Véronique Eglin"
      },
      {
        "literal": "Alexandre Aussem"
      },
      {
        "literal": "Jérémy Espinas"
      },
      {
        "literal": "Alexandre Lacoste"
      }
    ],
    "citation-key": "arxiv_2507_00210",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "b8972688fb95908c024a85e3ffc258be61280ee8d22e6ab97dfb0c4997b83445",
        "submitted_at": "2025-06-30T19:24:45Z",
        "updated_at": "2025-06-30T19:24:45Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2507.00210v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2507.00210",
      "catalog_id": "arxiv:2507.00210",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-06-30",
      "identifiers": {
        "arxiv": "2507.00210"
      },
      "record_updated_at": "2025-06-30"
    },
    "genre": "preprint",
    "id": "arxiv:2507.00210",
    "issued": {
      "date-parts": [
        [
          2025,
          6,
          30
        ]
      ]
    },
    "note": "First published: 2025-06-30; Recorded arXiv version: v1; Recorded version updated: 2025-06-30T19:24:45Z.",
    "title": "LineRetriever: Planning-Aware Observation Reduction for Web Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2507.05791v5",
    "archive": "arXiv",
    "archive_location": "2507.05791",
    "author": [
      {
        "literal": "Yan Yang"
      },
      {
        "literal": "Dongxu Li"
      },
      {
        "literal": "Yutong Dai"
      },
      {
        "literal": "Yuhao Yang"
      },
      {
        "literal": "Ziyang Luo"
      },
      {
        "literal": "Zirui Zhao"
      },
      {
        "literal": "Zhiyuan Hu"
      },
      {
        "literal": "Junzhe Huang"
      },
      {
        "literal": "Amrita Saha"
      },
      {
        "literal": "Zeyuan Chen"
      },
      {
        "literal": "Ran Xu"
      },
      {
        "literal": "Liyuan Pan"
      },
      {
        "literal": "Silvio Savarese"
      },
      {
        "literal": "Caiming Xiong"
      },
      {
        "literal": "Junnan Li"
      }
    ],
    "citation-key": "arxiv_2507_05791",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "3c1f659a30e1a0bd04219b9383927ce72e5f5b0d20783a89b850036d087155a4",
        "submitted_at": "2025-07-08T08:52:18Z",
        "updated_at": "2025-10-03T23:50:19Z",
        "version": 5,
        "versioned_url": "https://arxiv.org/abs/2507.05791v5"
      },
      "canonical_source_url": "https://arxiv.org/abs/2507.05791",
      "catalog_id": "arxiv:2507.05791",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-07-08",
      "identifiers": {
        "arxiv": "2507.05791"
      },
      "record_updated_at": "2025-10-03"
    },
    "genre": "preprint",
    "id": "arxiv:2507.05791",
    "issued": {
      "date-parts": [
        [
          2025,
          7,
          8
        ]
      ]
    },
    "note": "First published: 2025-07-08; Recorded arXiv version: v5; Recorded version updated: 2025-10-03T23:50:19Z.",
    "title": "GTA1: GUI Test-time Scaling Agent",
    "type": "article",
    "version": "5"
  },
  {
    "URL": "https://arxiv.org/abs/2507.18165v1",
    "archive": "arXiv",
    "archive_location": "2507.18165",
    "author": [
      {
        "literal": "Yuheng Zhao"
      },
      {
        "literal": "Xueli Shu"
      },
      {
        "literal": "Liwen Fan"
      },
      {
        "literal": "Lin Gao"
      },
      {
        "literal": "Yu Zhang"
      },
      {
        "literal": "Siming Chen"
      }
    ],
    "citation-key": "arxiv_2507_18165",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "e3c271f330c5bf27e2831efd96fb1d6f843bd79942791686a866df1b4812de2f",
        "submitted_at": "2025-07-24T08:02:35Z",
        "updated_at": "2025-07-24T08:02:35Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2507.18165v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2507.18165",
      "catalog_id": "arxiv:2507.18165",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2025-07-24",
      "identifiers": {
        "arxiv": "2507.18165"
      },
      "record_updated_at": "2025-07-24"
    },
    "genre": "preprint",
    "id": "arxiv:2507.18165",
    "issued": {
      "date-parts": [
        [
          2025,
          7,
          24
        ]
      ]
    },
    "note": "First published: 2025-07-24; Recorded arXiv version: v1; Recorded version updated: 2025-07-24T08:02:35Z.",
    "title": "ProactiveVA: Proactive Visual Analytics with LLM-Based UI Agent",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2507.19132v1",
    "archive": "arXiv",
    "archive_location": "2507.19132",
    "author": [
      {
        "literal": "Xuetian Chen"
      },
      {
        "literal": "Yinghao Chen"
      },
      {
        "literal": "Xinfeng Yuan"
      },
      {
        "literal": "Zhuo Peng"
      },
      {
        "literal": "Lu Chen"
      },
      {
        "literal": "Yuekeng Li"
      },
      {
        "literal": "Zhoujia Zhang"
      },
      {
        "literal": "Yingqian Huang"
      },
      {
        "literal": "Leyan Huang"
      },
      {
        "literal": "Jiaqing Liang"
      },
      {
        "literal": "Tianbao Xie"
      },
      {
        "literal": "Zhiyong Wu"
      },
      {
        "literal": "Qiushi Sun"
      },
      {
        "literal": "Biqing Qi"
      },
      {
        "literal": "Bowen Zhou"
      }
    ],
    "citation-key": "arxiv_2507_19132",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "fcd5a34bf6187f48da8e08842678e1e21cd558b2bcc7aab2611ceee7bc0d0179",
        "submitted_at": "2025-07-25T10:14:53Z",
        "updated_at": "2025-07-25T10:14:53Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2507.19132v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2507.19132",
      "catalog_id": "arxiv:2507.19132",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-07-25",
      "identifiers": {
        "arxiv": "2507.19132"
      },
      "record_updated_at": "2025-07-25"
    },
    "genre": "preprint",
    "id": "arxiv:2507.19132",
    "issued": {
      "date-parts": [
        [
          2025,
          7,
          25
        ]
      ]
    },
    "note": "First published: 2025-07-25; Recorded arXiv version: v1; Recorded version updated: 2025-07-25T10:14:53Z.",
    "title": "OS-MAP: How Far Can Computer-Using Agents Go in Breadth and Depth?",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2507.19478v1",
    "archive": "arXiv",
    "archive_location": "2507.19478",
    "author": [
      {
        "literal": "Xuehui Wang"
      },
      {
        "literal": "Zhenyu Wu"
      },
      {
        "literal": "JingJing Xie"
      },
      {
        "literal": "Zichen Ding"
      },
      {
        "literal": "Bowen Yang"
      },
      {
        "literal": "Zehao Li"
      },
      {
        "literal": "Zhaoyang Liu"
      },
      {
        "literal": "Qingyun Li"
      },
      {
        "literal": "Xuan Dong"
      },
      {
        "literal": "Zhe Chen"
      },
      {
        "literal": "Weiyun Wang"
      },
      {
        "literal": "Xiangyu Zhao"
      },
      {
        "literal": "Jixuan Chen"
      },
      {
        "literal": "Haodong Duan"
      },
      {
        "literal": "Tianbao Xie"
      },
      {
        "literal": "Chenyu Yang"
      },
      {
        "literal": "Shiqian Su"
      },
      {
        "literal": "Yue Yu"
      },
      {
        "literal": "Yuan Huang"
      },
      {
        "literal": "Yiqian Liu"
      },
      {
        "literal": "Xiao Zhang"
      },
      {
        "literal": "Yanting Zhang"
      },
      {
        "literal": "Xiangyu Yue"
      },
      {
        "literal": "Weijie Su"
      },
      {
        "literal": "Xizhou Zhu"
      },
      {
        "literal": "Wei Shen"
      },
      {
        "literal": "Jifeng Dai"
      },
      {
        "literal": "Wenhai Wang"
      }
    ],
    "citation-key": "arxiv_2507_19478",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "8b91f7bacbac917e374aa6f656dcfbcb2e220ab3227fa2cff295c63a1c005f9f",
        "submitted_at": "2025-07-25T17:59:26Z",
        "updated_at": "2025-07-25T17:59:26Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2507.19478v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2507.19478",
      "catalog_id": "arxiv:2507.19478",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-07-25",
      "identifiers": {
        "arxiv": "2507.19478"
      },
      "record_updated_at": "2025-07-25"
    },
    "genre": "preprint",
    "id": "arxiv:2507.19478",
    "issued": {
      "date-parts": [
        [
          2025,
          7,
          25
        ]
      ]
    },
    "note": "First published: 2025-07-25; Recorded arXiv version: v1; Recorded version updated: 2025-07-25T17:59:26Z.",
    "title": "MMBench-GUI: Hierarchical Multi-Platform Evaluation Framework for GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2507.21953v1",
    "archive": "arXiv",
    "archive_location": "2507.21953",
    "author": [
      {
        "literal": "Yi Kong"
      },
      {
        "literal": "Dianxi Shi"
      },
      {
        "literal": "Guoli Yang"
      },
      {
        "literal": "Zhang ke-di"
      },
      {
        "literal": "Chenlin Huang"
      },
      {
        "literal": "Xiaopeng Li"
      },
      {
        "literal": "Songchang Jin"
      }
    ],
    "citation-key": "arxiv_2507_21953",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "197dbb0e2734f30349260444fca80a8ac830a5d4000a677d6417f46509737475",
        "submitted_at": "2025-07-29T16:05:32Z",
        "updated_at": "2025-07-29T16:05:32Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2507.21953v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2507.21953",
      "catalog_id": "arxiv:2507.21953",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-07-29",
      "identifiers": {
        "arxiv": "2507.21953"
      },
      "record_updated_at": "2025-07-29"
    },
    "genre": "preprint",
    "id": "arxiv:2507.21953",
    "issued": {
      "date-parts": [
        [
          2025,
          7,
          29
        ]
      ]
    },
    "note": "First published: 2025-07-29; Recorded arXiv version: v1; Recorded version updated: 2025-07-29T16:05:32Z.",
    "title": "MapAgent: Trajectory-Constructed Memory-Augmented Planning for Mobile Task Automation",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2508.01330v4",
    "archive": "arXiv",
    "archive_location": "2508.01330",
    "author": [
      {
        "literal": "Zihan Zheng"
      },
      {
        "literal": "Tianle Cui"
      },
      {
        "literal": "Taoran Wang"
      },
      {
        "literal": "Fengtao Wang"
      },
      {
        "literal": "Jiahui Pan"
      },
      {
        "literal": "Lewei He"
      },
      {
        "literal": "Qianglong Chen"
      }
    ],
    "citation-key": "arxiv_2508_01330",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "57d89909534af90cbc10b97a992238bde3ae976ac6cc97778b4345635cef552c",
        "submitted_at": "2025-08-02T11:53:41Z",
        "updated_at": "2026-04-20T02:44:32Z",
        "version": 4,
        "versioned_url": "https://arxiv.org/abs/2508.01330v4"
      },
      "canonical_source_url": "https://arxiv.org/abs/2508.01330",
      "catalog_id": "arxiv:2508.01330",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-02",
      "identifiers": {
        "arxiv": "2508.01330"
      },
      "record_updated_at": "2026-04-20"
    },
    "genre": "preprint",
    "id": "arxiv:2508.01330",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          2
        ]
      ]
    },
    "note": "First published: 2025-08-02; Recorded arXiv version: v4; Recorded version updated: 2026-04-20T02:44:32Z.",
    "title": "NaturalGAIA: A Verifiable Benchmark and Hierarchical Framework for Long-Horizon GUI Tasks",
    "type": "article",
    "version": "4"
  },
  {
    "URL": "https://arxiv.org/abs/2508.01858v3",
    "archive": "arXiv",
    "archive_location": "2508.01858",
    "author": [
      {
        "literal": "Yuhan Guo"
      },
      {
        "literal": "Cong Guo"
      },
      {
        "literal": "Aiwen Sun"
      },
      {
        "literal": "Hongliang He"
      },
      {
        "literal": "Xinyu Yang"
      },
      {
        "literal": "Yue Lu"
      },
      {
        "literal": "Yingji Zhang"
      },
      {
        "literal": "Xuntao Guo"
      },
      {
        "literal": "Dong Zhang"
      },
      {
        "literal": "Jianzhuang Liu"
      },
      {
        "literal": "Jiang Duan"
      },
      {
        "literal": "Yijia Xiao"
      },
      {
        "literal": "Liangjian Wen"
      },
      {
        "literal": "Hai-Ming Xu"
      },
      {
        "literal": "Yong Dai"
      }
    ],
    "citation-key": "arxiv_2508_01858",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "7406b48ed9826d51215d955c31c3c67cd2ce9e08bf9a93a55ac4e85a546315f1",
        "submitted_at": "2025-08-03T17:17:52Z",
        "updated_at": "2026-07-05T06:17:25Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2508.01858v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2508.01858",
      "catalog_id": "arxiv:2508.01858",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-03",
      "identifiers": {
        "arxiv": "2508.01858"
      },
      "record_updated_at": "2026-07-05"
    },
    "genre": "preprint",
    "id": "arxiv:2508.01858",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          3
        ]
      ]
    },
    "note": "First published: 2025-08-03; Recorded arXiv version: v3; Recorded version updated: 2026-07-05T06:17:25Z.",
    "title": "Web-CogReasoner: Towards Multimodal Knowledge-Induced Cognitive Reasoning for Web Agents",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2508.03923v3",
    "archive": "arXiv",
    "archive_location": "2508.03923",
    "author": [
      {
        "literal": "Linxin Song"
      },
      {
        "literal": "Yutong Dai"
      },
      {
        "literal": "Viraj Prabhu"
      },
      {
        "literal": "Jieyu Zhang"
      },
      {
        "literal": "Taiwei Shi"
      },
      {
        "literal": "Li Li"
      },
      {
        "literal": "Junnan Li"
      },
      {
        "literal": "Silvio Savarese"
      },
      {
        "literal": "Zeyuan Chen"
      },
      {
        "literal": "Jieyu Zhao"
      },
      {
        "literal": "Ran Xu"
      },
      {
        "literal": "Caiming Xiong"
      }
    ],
    "citation-key": "arxiv_2508_03923",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "b6e4272a69c1a93f22bd35032547ffb6021a79c97d179a1075e14aee69dea83a",
        "submitted_at": "2025-08-05T21:33:36Z",
        "updated_at": "2026-02-20T10:14:08Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2508.03923v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2508.03923",
      "catalog_id": "arxiv:2508.03923",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-05",
      "identifiers": {
        "arxiv": "2508.03923"
      },
      "record_updated_at": "2026-02-20"
    },
    "genre": "preprint",
    "id": "arxiv:2508.03923",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          5
        ]
      ]
    },
    "note": "First published: 2025-08-05; Recorded arXiv version: v3; Recorded version updated: 2026-02-20T10:14:08Z.",
    "title": "CoAct-1: Computer-using Multi-Agent System with Coding Actions",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2508.04037v2",
    "archive": "arXiv",
    "archive_location": "2508.04037",
    "author": [
      {
        "literal": "Yuhao Cheng"
      },
      {
        "literal": "Liang Tang"
      },
      {
        "literal": "Shuxian Li"
      },
      {
        "literal": "Yukang Huo"
      },
      {
        "literal": "Tiaonan Duan"
      },
      {
        "literal": "Kaer Huang"
      },
      {
        "literal": "Yanzhe Jing"
      },
      {
        "literal": "Yiqiang Yan"
      }
    ],
    "citation-key": "arxiv_2508_04037",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d54ad74695a60cb040d88fcb85549912d7d4ecf6ca2ed78e636d05f8f17d1795",
        "submitted_at": "2025-08-06T02:57:22Z",
        "updated_at": "2026-01-22T05:22:03Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2508.04037v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2508.04037",
      "catalog_id": "arxiv:2508.04037",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-06",
      "identifiers": {
        "arxiv": "2508.04037"
      },
      "record_updated_at": "2026-01-22"
    },
    "genre": "preprint",
    "id": "arxiv:2508.04037",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          6
        ]
      ]
    },
    "note": "First published: 2025-08-06; Recorded arXiv version: v2; Recorded version updated: 2026-01-22T05:22:03Z.",
    "title": "Evolving in Tasks: Empowering the Multi-modality Large Language Model as the Computer Use Agent",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2508.05081v1",
    "archive": "arXiv",
    "archive_location": "2508.05081",
    "author": [
      {
        "literal": "Jiarun Liu"
      },
      {
        "literal": "Chunhong Zhang"
      },
      {
        "literal": "Zheng Hu"
      }
    ],
    "citation-key": "arxiv_2508_05081",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "2d574d4157b775639c08ed12813892584b011dd3c0271ca2586f9d6a61d00c14",
        "submitted_at": "2025-08-07T07:05:22Z",
        "updated_at": "2025-08-07T07:05:22Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2508.05081v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2508.05081",
      "catalog_id": "arxiv:2508.05081",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-07",
      "identifiers": {
        "arxiv": "2508.05081"
      },
      "record_updated_at": "2025-08-07"
    },
    "genre": "preprint",
    "id": "arxiv:2508.05081",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          7
        ]
      ]
    },
    "note": "First published: 2025-08-07; Recorded arXiv version: v1; Recorded version updated: 2025-08-07T07:05:22Z.",
    "title": "Cognitive Duality for Adaptive Web Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2508.09123v3",
    "archive": "arXiv",
    "archive_location": "2508.09123",
    "author": [
      {
        "literal": "Xinyuan Wang"
      },
      {
        "literal": "Bowen Wang"
      },
      {
        "literal": "Dunjie Lu"
      },
      {
        "literal": "Junlin Yang"
      },
      {
        "literal": "Tianbao Xie"
      },
      {
        "literal": "Junli Wang"
      },
      {
        "literal": "Jiaqi Deng"
      },
      {
        "literal": "Xiaole Guo"
      },
      {
        "literal": "Yiheng Xu"
      },
      {
        "literal": "Chen Henry Wu"
      },
      {
        "literal": "Zhennan Shen"
      },
      {
        "literal": "Zhuokai Li"
      },
      {
        "literal": "Ryan Li"
      },
      {
        "literal": "Xiaochuan Li"
      },
      {
        "literal": "Junda Chen"
      },
      {
        "literal": "Boyuan Zheng"
      },
      {
        "literal": "Peihang Li"
      },
      {
        "literal": "Fangyu Lei"
      },
      {
        "literal": "Ruisheng Cao"
      },
      {
        "literal": "Yeqiao Fu"
      },
      {
        "literal": "Dongchan Shin"
      },
      {
        "literal": "Martin Shin"
      },
      {
        "literal": "Jiarui Hu"
      },
      {
        "literal": "Yuyan Wang"
      },
      {
        "literal": "Jixuan Chen"
      },
      {
        "literal": "Yuxiao Ye"
      },
      {
        "literal": "Danyang Zhang"
      },
      {
        "literal": "Dikang Du"
      },
      {
        "literal": "Hao Hu"
      },
      {
        "literal": "Huarong Chen"
      },
      {
        "literal": "Zaida Zhou"
      },
      {
        "literal": "Haotian Yao"
      },
      {
        "literal": "Ziwei Chen"
      },
      {
        "literal": "Qizheng Gu"
      },
      {
        "literal": "Yipu Wang"
      },
      {
        "literal": "Heng Wang"
      },
      {
        "literal": "Diyi Yang"
      },
      {
        "literal": "Victor Zhong"
      },
      {
        "literal": "Flood Sung"
      },
      {
        "literal": "Y. Charles"
      },
      {
        "literal": "Zhilin Yang"
      },
      {
        "literal": "Tao Yu"
      }
    ],
    "citation-key": "arxiv_2508_09123",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "40e46f27128d066358fcee6279b26bb2b708dacaf1f7b5794b755520fe892243",
        "submitted_at": "2025-08-12T17:52:32Z",
        "updated_at": "2025-10-04T17:38:02Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2508.09123v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2508.09123",
      "catalog_id": "arxiv:2508.09123",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-12",
      "identifiers": {
        "arxiv": "2508.09123"
      },
      "record_updated_at": "2025-10-04"
    },
    "genre": "preprint",
    "id": "arxiv:2508.09123",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          12
        ]
      ]
    },
    "note": "First published: 2025-08-12; Recorded arXiv version: v3; Recorded version updated: 2025-10-04T17:38:02Z.",
    "title": "OpenCUA: Open Foundations for Computer-Use Agents",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2508.15144v2",
    "archive": "arXiv",
    "archive_location": "2508.15144",
    "author": [
      {
        "literal": "Jiabo Ye"
      },
      {
        "literal": "Xi Zhang"
      },
      {
        "literal": "Haiyang Xu"
      },
      {
        "literal": "Haowei Liu"
      },
      {
        "literal": "Junyang Wang"
      },
      {
        "literal": "Zhaoqing Zhu"
      },
      {
        "literal": "Ziwei Zheng"
      },
      {
        "literal": "Feiyu Gao"
      },
      {
        "literal": "Junjie Cao"
      },
      {
        "literal": "Zhengxi Lu"
      },
      {
        "literal": "Jitong Liao"
      },
      {
        "literal": "Qi Zheng"
      },
      {
        "literal": "Fei Huang"
      },
      {
        "literal": "Jingren Zhou"
      },
      {
        "literal": "Ming Yan"
      }
    ],
    "citation-key": "arxiv_2508_15144",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "629421d60bb8beeba3f087854ebaddc5461f9af0e5fb0bc572d0c20394ba9dcd",
        "submitted_at": "2025-08-21T00:39:12Z",
        "updated_at": "2025-09-01T05:38:34Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2508.15144v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2508.15144",
      "catalog_id": "arxiv:2508.15144",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-21",
      "identifiers": {
        "arxiv": "2508.15144"
      },
      "record_updated_at": "2025-09-01"
    },
    "genre": "preprint",
    "id": "arxiv:2508.15144",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          21
        ]
      ]
    },
    "note": "First published: 2025-08-21; Recorded arXiv version: v2; Recorded version updated: 2025-09-01T05:38:34Z.",
    "title": "Mobile-Agent-v3: Fundamental Agents for GUI Automation",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2508.16987v1",
    "archive": "arXiv",
    "archive_location": "2508.16987",
    "author": [
      {
        "literal": "Tanvir Bhathal"
      },
      {
        "literal": "Asanshay Gupta"
      }
    ],
    "citation-key": "arxiv_2508_16987",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "67da98d98fec647de9b482fabd8c7861063c0fdb91f1ec303fe8df759ee0f69d",
        "submitted_at": "2025-08-23T11:02:59Z",
        "updated_at": "2025-08-23T11:02:59Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2508.16987v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2508.16987",
      "catalog_id": "arxiv:2508.16987",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-23",
      "identifiers": {
        "arxiv": "2508.16987"
      },
      "record_updated_at": "2025-08-23"
    },
    "genre": "preprint",
    "id": "arxiv:2508.16987",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          23
        ]
      ]
    },
    "note": "First published: 2025-08-23; Recorded arXiv version: v1; Recorded version updated: 2025-08-23T11:02:59Z.",
    "title": "WebSight: A Vision-First Architecture for Robust Web Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2508.18040v1",
    "archive": "arXiv",
    "archive_location": "2508.18040",
    "author": [
      {
        "literal": "Xin Wang"
      },
      {
        "literal": "Zhiyao Cui"
      },
      {
        "literal": "Hao Li"
      },
      {
        "literal": "Ya Zeng"
      },
      {
        "literal": "Chenxu Wang"
      },
      {
        "literal": "Ruiqi Song"
      },
      {
        "literal": "Yihang Chen"
      },
      {
        "literal": "Kun Shao"
      },
      {
        "literal": "Qiaosheng Zhang"
      },
      {
        "literal": "Jinzhuo Liu"
      },
      {
        "literal": "Siyue Ren"
      },
      {
        "literal": "Shuyue Hu"
      },
      {
        "literal": "Zhen Wang"
      }
    ],
    "citation-key": "arxiv_2508_18040",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "57d5fc98a2eb2232f237d71b34060df0acef89cfeaa4bc247580a4dda3223e0b",
        "submitted_at": "2025-08-25T13:57:02Z",
        "updated_at": "2025-08-25T13:57:02Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2508.18040v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2508.18040",
      "catalog_id": "arxiv:2508.18040",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-25",
      "identifiers": {
        "arxiv": "2508.18040"
      },
      "record_updated_at": "2025-08-25"
    },
    "genre": "preprint",
    "id": "arxiv:2508.18040",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          25
        ]
      ]
    },
    "note": "First published: 2025-08-25; Recorded arXiv version: v1; Recorded version updated: 2025-08-25T13:57:02Z.",
    "title": "PerPilot: Personalizing VLM-based Mobile Agents via Memory and Exploration",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2508.18689v2",
    "archive": "arXiv",
    "archive_location": "2508.18689",
    "author": [
      {
        "literal": "Yuyang Zhao"
      },
      {
        "literal": "Wentao Shi"
      },
      {
        "literal": "Fuli Feng"
      },
      {
        "literal": "Xiangnan He"
      }
    ],
    "citation-key": "arxiv_2508_18689",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "02f4a76ae220caa289b322b622be6356086c5b04bc0f1598d0cd580c9ee25b6d",
        "submitted_at": "2025-08-26T05:23:24Z",
        "updated_at": "2025-08-27T04:25:35Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2508.18689v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2508.18689",
      "catalog_id": "arxiv:2508.18689",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-26",
      "identifiers": {
        "arxiv": "2508.18689"
      },
      "record_updated_at": "2025-08-27"
    },
    "genre": "preprint",
    "id": "arxiv:2508.18689",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          26
        ]
      ]
    },
    "note": "First published: 2025-08-26; Recorded arXiv version: v2; Recorded version updated: 2025-08-27T04:25:35Z.",
    "title": "AppAgent-Pro: A Proactive GUI Agent System for Multidomain Information Integration and User Assistance",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2508.20096v1",
    "archive": "arXiv",
    "archive_location": "2508.20096",
    "author": [
      {
        "literal": "Zeyi Sun"
      },
      {
        "literal": "Yuhang Cao"
      },
      {
        "literal": "Jianze Liang"
      },
      {
        "literal": "Qiushi Sun"
      },
      {
        "literal": "Ziyu Liu"
      },
      {
        "literal": "Zhixiong Zhang"
      },
      {
        "literal": "Yuhang Zang"
      },
      {
        "literal": "Xiaoyi Dong"
      },
      {
        "literal": "Kai Chen"
      },
      {
        "literal": "Dahua Lin"
      },
      {
        "literal": "Jiaqi Wang"
      }
    ],
    "citation-key": "arxiv_2508_20096",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "ac401f8d5c56050e72839f3bffba4f2c78bbdf34a00cab05fe504f03efd96a19",
        "submitted_at": "2025-08-27T17:59:50Z",
        "updated_at": "2025-08-27T17:59:50Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2508.20096v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2508.20096",
      "catalog_id": "arxiv:2508.20096",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-27",
      "identifiers": {
        "arxiv": "2508.20096"
      },
      "record_updated_at": "2025-08-27"
    },
    "genre": "preprint",
    "id": "arxiv:2508.20096",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          27
        ]
      ]
    },
    "note": "First published: 2025-08-27; Recorded arXiv version: v1; Recorded version updated: 2025-08-27T17:59:50Z.",
    "title": "CODA: Coordinating the Cerebrum and Cerebellum for a Dual-Brain Computer Use Agent with Decoupled Reinforcement Learning",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2508.21767v1",
    "archive": "arXiv",
    "archive_location": "2508.21767",
    "author": [
      {
        "literal": "Zhixiong Zeng"
      },
      {
        "literal": "Jing Huang"
      },
      {
        "literal": "Liming Zheng"
      },
      {
        "literal": "Wenkang Han"
      },
      {
        "literal": "Yufeng Zhong"
      },
      {
        "literal": "Lei Chen"
      },
      {
        "literal": "Longrong Yang"
      },
      {
        "literal": "Yingjie Chu"
      },
      {
        "literal": "Yuzhi He"
      },
      {
        "literal": "Lin Ma"
      }
    ],
    "citation-key": "arxiv_2508_21767",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "b2a14819c7d7e8258947c9db8cc89012f4d133eaced4e4b904c3bfd1761181e5",
        "submitted_at": "2025-08-29T16:40:57Z",
        "updated_at": "2025-08-29T16:40:57Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2508.21767v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2508.21767",
      "catalog_id": "arxiv:2508.21767",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-29",
      "identifiers": {
        "arxiv": "2508.21767"
      },
      "record_updated_at": "2025-08-29"
    },
    "genre": "preprint",
    "id": "arxiv:2508.21767",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          29
        ]
      ]
    },
    "note": "First published: 2025-08-29; Recorded arXiv version: v1; Recorded version updated: 2025-08-29T16:40:57Z.",
    "title": "UItron: Foundational GUI Agent with Advanced Perception and Planning",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2509.00366v1",
    "archive": "arXiv",
    "archive_location": "2509.00366",
    "author": [
      {
        "literal": "Ziyi Guan"
      },
      {
        "literal": "Jason Chun Lok Li"
      },
      {
        "literal": "Zhijian Hou"
      },
      {
        "literal": "Pingping Zhang"
      },
      {
        "literal": "Donglai Xu"
      },
      {
        "literal": "Yuzhi Zhao"
      },
      {
        "literal": "Mengyang Wu"
      },
      {
        "literal": "Jinpeng Chen"
      },
      {
        "literal": "Thanh-Toan Nguyen"
      },
      {
        "literal": "Pengfei Xian"
      },
      {
        "literal": "Wenao Ma"
      },
      {
        "literal": "Shengchao Qin"
      },
      {
        "literal": "Graziano Chesi"
      },
      {
        "literal": "Ngai Wong"
      }
    ],
    "citation-key": "arxiv_2509_00366",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "c22f63bb5de22dd922ba37ea45df9404def017fee492e8ade03c83ab8b4f9b58",
        "submitted_at": "2025-08-30T05:32:32Z",
        "updated_at": "2025-08-30T05:32:32Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2509.00366v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2509.00366",
      "catalog_id": "arxiv:2509.00366",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-08-30",
      "identifiers": {
        "arxiv": "2509.00366"
      },
      "record_updated_at": "2025-08-30"
    },
    "genre": "preprint",
    "id": "arxiv:2509.00366",
    "issued": {
      "date-parts": [
        [
          2025,
          8,
          30
        ]
      ]
    },
    "note": "First published: 2025-08-30; Recorded arXiv version: v1; Recorded version updated: 2025-08-30T05:32:32Z.",
    "title": "KG-RAG: Enhancing GUI Agent Decision-Making via Knowledge Graph-Driven Retrieval-Augmented Generation",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2509.01052v2",
    "archive": "arXiv",
    "archive_location": "2509.01052",
    "author": [
      {
        "literal": "Jaewoo Ahn"
      },
      {
        "literal": "Junseo Kim"
      },
      {
        "literal": "Heeseung Yun"
      },
      {
        "literal": "Jaehyeon Son"
      },
      {
        "literal": "Dongmin Park"
      },
      {
        "literal": "Jaewoong Cho"
      },
      {
        "literal": "Gunhee Kim"
      }
    ],
    "citation-key": "arxiv_2509_01052",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "3bf86652aaeb252972f3010be38b75b78dfaa9a8063c480c80d0cea47397c2e7",
        "submitted_at": "2025-09-01T01:33:16Z",
        "updated_at": "2025-10-15T10:33:27Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2509.01052v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2509.01052",
      "catalog_id": "arxiv:2509.01052",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2025-09-01",
      "identifiers": {
        "arxiv": "2509.01052"
      },
      "record_updated_at": "2025-10-15"
    },
    "genre": "preprint",
    "id": "arxiv:2509.01052",
    "issued": {
      "date-parts": [
        [
          2025,
          9,
          1
        ]
      ]
    },
    "note": "First published: 2025-09-01; Recorded arXiv version: v2; Recorded version updated: 2025-10-15T10:33:27Z.",
    "title": "FlashAdventure: A Benchmark for GUI Agents Solving Full Story Arcs in Diverse Adventure Games",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2509.02444v2",
    "archive": "arXiv",
    "archive_location": "2509.02444",
    "author": [
      {
        "literal": "Jingru Fan"
      },
      {
        "literal": "Yufan Dang"
      },
      {
        "literal": "Jingyao Wu"
      },
      {
        "literal": "Huatao Li"
      },
      {
        "literal": "Runde Yang"
      },
      {
        "literal": "Xiyuan Yang"
      },
      {
        "literal": "Yuheng Wang"
      },
      {
        "literal": "Chen Qian"
      }
    ],
    "citation-key": "arxiv_2509_02444",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "49a887b12b3395821aa100a66e61aac808309ef947ae752475df3d2db9a4fcd8",
        "submitted_at": "2025-09-02T15:48:21Z",
        "updated_at": "2025-10-17T00:57:58Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2509.02444v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2509.02444",
      "catalog_id": "arxiv:2509.02444",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-09-02",
      "identifiers": {
        "arxiv": "2509.02444"
      },
      "record_updated_at": "2025-10-17"
    },
    "genre": "preprint",
    "id": "arxiv:2509.02444",
    "issued": {
      "date-parts": [
        [
          2025,
          9,
          2
        ]
      ]
    },
    "note": "First published: 2025-09-02; Recorded arXiv version: v2; Recorded version updated: 2025-10-17T00:57:58Z.",
    "title": "AppCopilot: Toward General, Accurate, Long-Horizon, and Efficient Mobile Agent",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2509.13704v1",
    "archive": "arXiv",
    "archive_location": "2509.13704",
    "author": [
      {
        "literal": "Liangtao Lin"
      },
      {
        "literal": "Zhaomeng Zhu"
      },
      {
        "literal": "Tianwei Zhang"
      },
      {
        "literal": "Yonggang Wen"
      }
    ],
    "citation-key": "arxiv_2509_13704",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "2c1a4019df77278943829e16b155771abb1c8ca30726044cf5969227e0437f5f",
        "submitted_at": "2025-09-17T05:14:11Z",
        "updated_at": "2025-09-17T05:14:11Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2509.13704v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2509.13704",
      "catalog_id": "arxiv:2509.13704",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-09-17",
      "identifiers": {
        "arxiv": "2509.13704"
      },
      "record_updated_at": "2025-09-17"
    },
    "genre": "preprint",
    "id": "arxiv:2509.13704",
    "issued": {
      "date-parts": [
        [
          2025,
          9,
          17
        ]
      ]
    },
    "note": "First published: 2025-09-17; Recorded arXiv version: v1; Recorded version updated: 2025-09-17T05:14:11Z.",
    "title": "InfraMind: A Novel Exploration-based GUI Agentic Framework for Mission-critical Industrial Management",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2509.15566v4",
    "archive": "arXiv",
    "archive_location": "2509.15566",
    "author": [
      {
        "literal": "Shaojie Zhang"
      },
      {
        "literal": "Ruoceng Zhang"
      },
      {
        "literal": "Pei Fu"
      },
      {
        "literal": "Shaokang Wang"
      },
      {
        "literal": "Jiahui Yang"
      },
      {
        "literal": "Xin Du"
      },
      {
        "literal": "Shiqi Cui"
      },
      {
        "literal": "Bin Qin"
      },
      {
        "literal": "Ying Huang"
      },
      {
        "literal": "Zhenbo Luo"
      },
      {
        "literal": "Jian Luan"
      }
    ],
    "citation-key": "arxiv_2509_15566",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d190eefae9309d87f4ea4780d3483ae0ebd834e0bf70e7f2af204c3788d87e7e",
        "submitted_at": "2025-09-19T04:03:44Z",
        "updated_at": "2025-10-27T11:34:58Z",
        "version": 4,
        "versioned_url": "https://arxiv.org/abs/2509.15566v4"
      },
      "canonical_source_url": "https://arxiv.org/abs/2509.15566",
      "catalog_id": "arxiv:2509.15566",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-09-19",
      "identifiers": {
        "arxiv": "2509.15566"
      },
      "record_updated_at": "2025-10-27"
    },
    "genre": "preprint",
    "id": "arxiv:2509.15566",
    "issued": {
      "date-parts": [
        [
          2025,
          9,
          19
        ]
      ]
    },
    "note": "First published: 2025-09-19; Recorded arXiv version: v4; Recorded version updated: 2025-10-27T11:34:58Z.",
    "title": "BTL-UI: Blink-Think-Link Reasoning Model for GUI Agent",
    "type": "article",
    "version": "4"
  },
  {
    "URL": "https://arxiv.org/abs/2509.15738v1",
    "archive": "arXiv",
    "archive_location": "2509.15738",
    "author": [
      {
        "literal": "Musen Lin"
      },
      {
        "literal": "Minghao Liu"
      },
      {
        "literal": "Taoran Lu"
      },
      {
        "literal": "Lichen Yuan"
      },
      {
        "literal": "Yiwei Liu"
      },
      {
        "literal": "Haonan Xu"
      },
      {
        "literal": "Yu Miao"
      },
      {
        "literal": "Yuhao Chao"
      },
      {
        "literal": "Zhaojian Li"
      }
    ],
    "citation-key": "arxiv_2509_15738",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "035048459d2b303721ef0738212829d1fa3314ccd0736b6649cc786c6e975ad4",
        "submitted_at": "2025-09-19T08:09:18Z",
        "updated_at": "2025-09-19T08:09:18Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2509.15738v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2509.15738",
      "catalog_id": "arxiv:2509.15738",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-09-19",
      "identifiers": {
        "arxiv": "2509.15738"
      },
      "record_updated_at": "2025-09-19"
    },
    "genre": "preprint",
    "id": "arxiv:2509.15738",
    "issued": {
      "date-parts": [
        [
          2025,
          9,
          19
        ]
      ]
    },
    "note": "First published: 2025-09-19; Recorded arXiv version: v1; Recorded version updated: 2025-09-19T08:09:18Z.",
    "title": "GUI-ReWalk: Massive Data Generation for GUI Agent via Stochastic Exploration and Intent-Aware Reasoning",
    "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/2509.24133v1",
    "archive": "arXiv",
    "archive_location": "2509.24133",
    "author": [
      {
        "literal": "Zhecheng Li"
      },
      {
        "literal": "Guoxian Song"
      },
      {
        "literal": "Yiwei Wang"
      },
      {
        "literal": "Zhen Xiong"
      },
      {
        "literal": "Junsong Yuan"
      },
      {
        "literal": "Yujun Cai"
      }
    ],
    "citation-key": "arxiv_2509_24133",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "9b42e784de8dd007c60d234b7331bd98b5a04c58c13696bf436c81c7eccee37f",
        "submitted_at": "2025-09-29T00:06:31Z",
        "updated_at": "2025-09-29T00:06:31Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2509.24133v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2509.24133",
      "catalog_id": "arxiv:2509.24133",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-09-29",
      "identifiers": {
        "arxiv": "2509.24133"
      },
      "record_updated_at": "2025-09-29"
    },
    "genre": "preprint",
    "id": "arxiv:2509.24133",
    "issued": {
      "date-parts": [
        [
          2025,
          9,
          29
        ]
      ]
    },
    "note": "First published: 2025-09-29; Recorded arXiv version: v1; Recorded version updated: 2025-09-29T00:06:31Z.",
    "title": "Generalist Scanner Meets Specialist Locator: A Synergistic Coarse-to-Fine Framework for Robust GUI Grounding",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2509.24183v1",
    "archive": "arXiv",
    "archive_location": "2509.24183",
    "author": [
      {
        "literal": "Ran Xu"
      },
      {
        "literal": "Kaixin Ma"
      },
      {
        "literal": "Wenhao Yu"
      },
      {
        "literal": "Hongming Zhang"
      },
      {
        "literal": "Joyce C. Ho"
      },
      {
        "literal": "Carl Yang"
      },
      {
        "literal": "Dong Yu"
      }
    ],
    "citation-key": "arxiv_2509_24183",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "e1c6fd3e84212bac4c4c8341ffb3478c97becce08f92dac89e5de77ad42cd07d",
        "submitted_at": "2025-09-29T02:04:20Z",
        "updated_at": "2025-09-29T02:04:20Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2509.24183v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2509.24183",
      "catalog_id": "arxiv:2509.24183",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-09-29",
      "identifiers": {
        "arxiv": "2509.24183"
      },
      "record_updated_at": "2025-09-29"
    },
    "genre": "preprint",
    "id": "arxiv:2509.24183",
    "issued": {
      "date-parts": [
        [
          2025,
          9,
          29
        ]
      ]
    },
    "note": "First published: 2025-09-29; Recorded arXiv version: v1; Recorded version updated: 2025-09-29T02:04:20Z.",
    "title": "Retrieval-augmented GUI Agents with Generative Guidelines",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2510.00413v2",
    "archive": "arXiv",
    "archive_location": "2510.00413",
    "author": [
      {
        "literal": "Zikang Liu"
      },
      {
        "literal": "Junyi Li"
      },
      {
        "literal": "Wayne Xin Zhao"
      },
      {
        "literal": "Dawei Gao"
      },
      {
        "literal": "Yaliang Li"
      },
      {
        "literal": "Ji-rong Wen"
      }
    ],
    "citation-key": "arxiv_2510_00413",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "277315ecc1a341c0a511bce84b87190fc0aeeb2306666ea8d4746aa8e9b39984",
        "submitted_at": "2025-10-01T01:48:39Z",
        "updated_at": "2025-10-04T15:23:55Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2510.00413v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2510.00413",
      "catalog_id": "arxiv:2510.00413",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-10-01",
      "identifiers": {
        "arxiv": "2510.00413"
      },
      "record_updated_at": "2025-10-04"
    },
    "genre": "preprint",
    "id": "arxiv:2510.00413",
    "issued": {
      "date-parts": [
        [
          2025,
          10,
          1
        ]
      ]
    },
    "note": "First published: 2025-10-01; Recorded arXiv version: v2; Recorded version updated: 2025-10-04T15:23:55Z.",
    "title": "PAL-UI: Planning with Active Look-back for Vision-Based GUI Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2510.04039v1",
    "archive": "arXiv",
    "archive_location": "2510.04039",
    "author": [
      {
        "literal": "Bin Lei"
      },
      {
        "literal": "Nuo Xu"
      },
      {
        "literal": "Ali Payani"
      },
      {
        "literal": "Mingyi Hong"
      },
      {
        "literal": "Chunhua Liao"
      },
      {
        "literal": "Yu Cao"
      },
      {
        "literal": "Caiwen Ding"
      }
    ],
    "citation-key": "arxiv_2510_04039",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "798fc5ca4fe483a1235f0979b0c641e9c1732660780c13340c5221319a1b28fc",
        "submitted_at": "2025-10-05T05:15:45Z",
        "updated_at": "2025-10-05T05:15:45Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2510.04039v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2510.04039",
      "catalog_id": "arxiv:2510.04039",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-10-05",
      "identifiers": {
        "arxiv": "2510.04039"
      },
      "record_updated_at": "2025-10-05"
    },
    "genre": "preprint",
    "id": "arxiv:2510.04039",
    "issued": {
      "date-parts": [
        [
          2025,
          10,
          5
        ]
      ]
    },
    "note": "First published: 2025-10-05; Recorded arXiv version: v1; Recorded version updated: 2025-10-05T05:15:45Z.",
    "title": "\\textsc{GUI-Spotlight}: Adaptive Iterative Focus Refinement for Enhanced GUI Visual Grounding",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2510.04607v2",
    "archive": "arXiv",
    "archive_location": "2510.04607",
    "author": [
      {
        "literal": "Yuan Wang"
      },
      {
        "literal": "Mingyu Li"
      },
      {
        "literal": "Haibo Chen"
      }
    ],
    "citation-key": "arxiv_2510_04607",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "1c22547bf82eb2fb87e8bf6a3846d25c17cac296763726ed9a25faf88bdca33c",
        "submitted_at": "2025-10-06T09:14:58Z",
        "updated_at": "2026-03-25T11:29:08Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2510.04607v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2510.04607",
      "catalog_id": "arxiv:2510.04607",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-10-06",
      "identifiers": {
        "arxiv": "2510.04607"
      },
      "record_updated_at": "2026-03-25"
    },
    "genre": "preprint",
    "id": "arxiv:2510.04607",
    "issued": {
      "date-parts": [
        [
          2025,
          10,
          6
        ]
      ]
    },
    "note": "First published: 2025-10-06; Recorded arXiv version: v2; Recorded version updated: 2026-03-25T11:29:08Z.",
    "title": "From Imperative to Declarative: Towards LLM-friendly OS Interfaces for Boosted Computer-Use Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2510.06042v3",
    "archive": "arXiv",
    "archive_location": "2510.06042",
    "author": [
      {
        "literal": "Shang Ma"
      },
      {
        "literal": "Xusheng Xiao"
      },
      {
        "literal": "Yanfang Ye"
      }
    ],
    "citation-key": "arxiv_2510_06042",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "4bf0a561460c080b14da6fd13f6d92781e01e52b08dc691fe019e70b88e44c86",
        "submitted_at": "2025-10-07T15:36:04Z",
        "updated_at": "2026-09-27T17:25:57Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2510.06042v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2510.06042",
      "catalog_id": "arxiv:2510.06042",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-10-07",
      "identifiers": {
        "arxiv": "2510.06042"
      },
      "record_updated_at": "2026-09-27"
    },
    "genre": "preprint",
    "id": "arxiv:2510.06042",
    "issued": {
      "date-parts": [
        [
          2025,
          10,
          7
        ]
      ]
    },
    "note": "First published: 2025-10-07; Recorded arXiv version: v3; Recorded version updated: 2026-09-27T17:25:57Z.",
    "title": "Agent+P: Guiding UI Agents via Symbolic Planning",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2510.11892v2",
    "archive": "arXiv",
    "archive_location": "2510.11892",
    "author": [
      {
        "literal": "Kai Mei"
      },
      {
        "literal": "Jiang Guo"
      },
      {
        "literal": "Shuaichen Chang"
      },
      {
        "literal": "Mingwen Dong"
      },
      {
        "literal": "Dongkyu Lee"
      },
      {
        "literal": "Xing Niu"
      },
      {
        "literal": "Jiarong Jiang"
      }
    ],
    "citation-key": "arxiv_2510_11892",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "7c37dae372655e8ef556163744e47ed42db3533175ca7d1db7560033774ed2a0",
        "submitted_at": "2025-10-13T19:52:04Z",
        "updated_at": "2026-03-09T15:25:16Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2510.11892v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2510.11892",
      "catalog_id": "arxiv:2510.11892",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-10-13",
      "identifiers": {
        "arxiv": "2510.11892"
      },
      "record_updated_at": "2026-03-09"
    },
    "genre": "preprint",
    "id": "arxiv:2510.11892",
    "issued": {
      "date-parts": [
        [
          2025,
          10,
          13
        ]
      ]
    },
    "note": "First published: 2025-10-13; Recorded arXiv version: v2; Recorded version updated: 2026-03-09T15:25:16Z.",
    "title": "R-WoM: Retrieval-augmented World Model For Computer-use Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2510.14388v2",
    "archive": "arXiv",
    "archive_location": "2510.14388",
    "author": [
      {
        "literal": "Zhe Wu"
      },
      {
        "literal": "Hongjin Lu"
      },
      {
        "literal": "Junliang Xing"
      },
      {
        "literal": "Changhao Zhang"
      },
      {
        "literal": "Yuxuan Li"
      },
      {
        "literal": "Yin Zhu"
      },
      {
        "literal": "Yuhao Yang"
      },
      {
        "literal": "Yuheng Jing"
      },
      {
        "literal": "Kai Li"
      },
      {
        "literal": "Kun Shao"
      },
      {
        "literal": "Jianye Hao"
      },
      {
        "literal": "Jun Wang"
      },
      {
        "literal": "Yuanchun Shi"
      }
    ],
    "citation-key": "arxiv_2510_14388",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "95fd3e1286ed49d569b05644d77f822ea9747cfb4769cb7cb93f39c9271cf57e",
        "submitted_at": "2025-10-16T07:38:21Z",
        "updated_at": "2026-02-06T02:04:25Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2510.14388v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2510.14388",
      "catalog_id": "arxiv:2510.14388",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-10-16",
      "identifiers": {
        "arxiv": "2510.14388"
      },
      "record_updated_at": "2026-02-06"
    },
    "genre": "preprint",
    "id": "arxiv:2510.14388",
    "issued": {
      "date-parts": [
        [
          2025,
          10,
          16
        ]
      ]
    },
    "note": "First published: 2025-10-16; Recorded arXiv version: v2; Recorded version updated: 2026-02-06T02:04:25Z.",
    "title": "Hi-Agent: Hierarchical Vision-Language Agents for Mobile Device Control",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2510.19949v2",
    "archive": "arXiv",
    "archive_location": "2510.19949",
    "author": [
      {
        "literal": "Mathieu Andreux"
      },
      {
        "literal": "Märt Bakler"
      },
      {
        "literal": "Yanael Barbier"
      },
      {
        "literal": "Hamza Benchekroun"
      },
      {
        "literal": "Emilien Biré"
      },
      {
        "literal": "Antoine Bonnet"
      },
      {
        "literal": "Riaz Bordie"
      },
      {
        "literal": "Nathan Bout"
      },
      {
        "literal": "Matthias Brunel"
      },
      {
        "literal": "Aleix Cambray"
      },
      {
        "literal": "Pierre-Louis Cedoz"
      },
      {
        "literal": "Antoine Chassang"
      },
      {
        "literal": "Gautier Cloix"
      },
      {
        "literal": "Ethan Connelly"
      },
      {
        "literal": "Alexandra Constantinou"
      },
      {
        "literal": "Ramzi De Coster"
      },
      {
        "literal": "Hubert de la Jonquiere"
      },
      {
        "literal": "Aurélien Delfosse"
      },
      {
        "literal": "Maxime Delpit"
      },
      {
        "literal": "Alexis Deprez"
      },
      {
        "literal": "Augustin Derupti"
      },
      {
        "literal": "Mathieu Diaz"
      },
      {
        "literal": "Shannon D'Souza"
      },
      {
        "literal": "Julie Dujardin"
      },
      {
        "literal": "Abai Edmund"
      },
      {
        "literal": "Michael Eickenberg"
      },
      {
        "literal": "Armand Fatalot"
      },
      {
        "literal": "Wissem Felissi"
      },
      {
        "literal": "Isaac Herring"
      },
      {
        "literal": "Xavier Koegler"
      },
      {
        "literal": "Erwan Le Jumeau de Kergaradec"
      },
      {
        "literal": "Aurélien Lac"
      },
      {
        "literal": "Maxime Langevin"
      },
      {
        "literal": "Corentin Lauverjat"
      },
      {
        "literal": "Antonio Loison"
      },
      {
        "literal": "Avshalom Manevich"
      },
      {
        "literal": "Axel Moyal"
      },
      {
        "literal": "Axel Nguyen Kerbel"
      },
      {
        "literal": "Marinela Parovic"
      },
      {
        "literal": "Julien Revelle"
      },
      {
        "literal": "Guillaume Richard"
      },
      {
        "literal": "Mats Richter"
      },
      {
        "literal": "Ronan Riochet"
      },
      {
        "literal": "María Santos"
      },
      {
        "literal": "Romain Savidan"
      },
      {
        "literal": "Laurent Sifre"
      },
      {
        "literal": "Maxime Theillard"
      },
      {
        "literal": "Marc Thibault"
      },
      {
        "literal": "Ivan Valentini"
      },
      {
        "literal": "Tony Wu"
      },
      {
        "literal": "Laura Yie"
      },
      {
        "literal": "Kai Yuan"
      },
      {
        "literal": "Jevgenij Zubovskij"
      }
    ],
    "citation-key": "arxiv_2510_19949",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "de25494af1bc2ceaa760a21da51803aaf89d9d2d1961887cfd0023fb87f28f63",
        "submitted_at": "2025-10-22T18:21:52Z",
        "updated_at": "2025-10-24T11:52:29Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2510.19949v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2510.19949",
      "catalog_id": "arxiv:2510.19949",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-10-22",
      "identifiers": {
        "arxiv": "2510.19949"
      },
      "record_updated_at": "2025-10-24"
    },
    "genre": "preprint",
    "id": "arxiv:2510.19949",
    "issued": {
      "date-parts": [
        [
          2025,
          10,
          22
        ]
      ]
    },
    "note": "First published: 2025-10-22; Recorded arXiv version: v2; Recorded version updated: 2025-10-24T11:52:29Z.",
    "title": "Surfer 2: The Next Generation of Cross-Platform Computer Use Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2510.24563v2",
    "archive": "arXiv",
    "archive_location": "2510.24563",
    "author": [
      {
        "literal": "Hongrui Jia"
      },
      {
        "literal": "Jitong Liao"
      },
      {
        "literal": "Xi Zhang"
      },
      {
        "literal": "Haiyang Xu"
      },
      {
        "literal": "Tianbao Xie"
      },
      {
        "literal": "Chaoya Jiang"
      },
      {
        "literal": "Ming Yan"
      },
      {
        "literal": "Si Liu"
      },
      {
        "literal": "Wei Ye"
      },
      {
        "literal": "Fei Huang"
      }
    ],
    "citation-key": "arxiv_2510_24563",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d115356f723746ce95c3e40fed933786a0747b3b560c8f73a6b7af05ca22d24f",
        "submitted_at": "2025-10-28T15:56:36Z",
        "updated_at": "2025-11-11T04:10:32Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2510.24563v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2510.24563",
      "catalog_id": "arxiv:2510.24563",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-10-28",
      "identifiers": {
        "arxiv": "2510.24563"
      },
      "record_updated_at": "2025-11-11"
    },
    "genre": "preprint",
    "id": "arxiv:2510.24563",
    "issued": {
      "date-parts": [
        [
          2025,
          10,
          28
        ]
      ]
    },
    "note": "First published: 2025-10-28; Recorded arXiv version: v2; Recorded version updated: 2025-11-11T04:10:32Z.",
    "title": "OSWorld-MCP: Benchmarking MCP Tool Invocation In Computer-Use Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2510.27210v1",
    "archive": "arXiv",
    "archive_location": "2510.27210",
    "author": [
      {
        "literal": "Tao Liu"
      },
      {
        "literal": "Chongyu Wang"
      },
      {
        "literal": "Rongjie Li"
      },
      {
        "literal": "Yingchen Yu"
      },
      {
        "literal": "Xuming He"
      },
      {
        "literal": "Bai Song"
      }
    ],
    "citation-key": "arxiv_2510_27210",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "283b1a1de2a4e1d4ec8175f0c15af66c10097a7f5ad9fd2793e93b888526ff7e",
        "submitted_at": "2025-10-31T06:10:57Z",
        "updated_at": "2025-10-31T06:10:57Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2510.27210v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2510.27210",
      "catalog_id": "arxiv:2510.27210",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-10-31",
      "identifiers": {
        "arxiv": "2510.27210"
      },
      "record_updated_at": "2025-10-31"
    },
    "genre": "preprint",
    "id": "arxiv:2510.27210",
    "issued": {
      "date-parts": [
        [
          2025,
          10,
          31
        ]
      ]
    },
    "note": "First published: 2025-10-31; Recorded arXiv version: v1; Recorded version updated: 2025-10-31T06:10:57Z.",
    "title": "GUI-Rise: Structured Reasoning and History Summarization for GUI Navigation",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2511.04137v1",
    "archive": "arXiv",
    "archive_location": "2511.04137",
    "author": [
      {
        "literal": "Yujian Liu"
      },
      {
        "literal": "Ze Wang"
      },
      {
        "literal": "Hao Chen"
      },
      {
        "literal": "Ximeng Sun"
      },
      {
        "literal": "Xiaodong Yu"
      },
      {
        "literal": "Jialian Wu"
      },
      {
        "literal": "Jiang Liu"
      },
      {
        "literal": "Emad Barsoum"
      },
      {
        "literal": "Zicheng Liu"
      },
      {
        "literal": "Shiyu Chang"
      }
    ],
    "citation-key": "arxiv_2511_04137",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "a91413c1e80869860d411402ab49f6e5968787b3bca83f0a9066a0e9c2f75d40",
        "submitted_at": "2025-11-06T07:29:02Z",
        "updated_at": "2025-11-06T07:29:02Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2511.04137v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2511.04137",
      "catalog_id": "arxiv:2511.04137",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-11-06",
      "identifiers": {
        "arxiv": "2511.04137"
      },
      "record_updated_at": "2025-11-06"
    },
    "genre": "preprint",
    "id": "arxiv:2511.04137",
    "issued": {
      "date-parts": [
        [
          2025,
          11,
          6
        ]
      ]
    },
    "note": "First published: 2025-11-06; Recorded arXiv version: v1; Recorded version updated: 2025-11-06T07:29:02Z.",
    "title": "Learning from Online Videos at Inference Time for Computer-Use Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2511.10705v1",
    "archive": "arXiv",
    "archive_location": "2511.10705",
    "author": [
      {
        "literal": "Yuan Zhao"
      },
      {
        "literal": "Hualei Zhu"
      },
      {
        "literal": "Tingyu Jiang"
      },
      {
        "literal": "Shen Li"
      },
      {
        "literal": "Xiaohang Xu"
      },
      {
        "literal": "Hao Henry Wang"
      }
    ],
    "citation-key": "arxiv_2511_10705",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "101d38ed439c41ebfa826e990697b502b9126728956295d25b0ea9d539ff23e3",
        "submitted_at": "2025-11-13T03:41:02Z",
        "updated_at": "2025-11-13T03:41:02Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2511.10705v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2511.10705",
      "catalog_id": "arxiv:2511.10705",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-11-13",
      "identifiers": {
        "arxiv": "2511.10705"
      },
      "record_updated_at": "2025-11-13"
    },
    "genre": "preprint",
    "id": "arxiv:2511.10705",
    "issued": {
      "date-parts": [
        [
          2025,
          11,
          13
        ]
      ]
    },
    "note": "First published: 2025-11-13; Recorded arXiv version: v1; Recorded version updated: 2025-11-13T03:41:02Z.",
    "title": "Co-EPG: A Framework for Co-Evolution of Planning and Grounding in Autonomous GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2511.22235v2",
    "archive": "arXiv",
    "archive_location": "2511.22235",
    "author": [
      {
        "literal": "Zehao Deng"
      },
      {
        "literal": "Tianjie Ju"
      },
      {
        "literal": "Zheng Wu"
      },
      {
        "literal": "Zhuosheng Zhang"
      },
      {
        "literal": "Gongshen Liu"
      }
    ],
    "citation-key": "arxiv_2511_22235",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "9acb395dfaad18b84ea4b52d97be194a42eb42d6ac922c55f2179e38cc258a31",
        "submitted_at": "2025-11-27T09:01:38Z",
        "updated_at": "2026-03-04T12:15:06Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2511.22235v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2511.22235",
      "catalog_id": "arxiv:2511.22235",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-11-27",
      "identifiers": {
        "arxiv": "2511.22235"
      },
      "record_updated_at": "2026-03-04"
    },
    "genre": "preprint",
    "id": "arxiv:2511.22235",
    "issued": {
      "date-parts": [
        [
          2025,
          11,
          27
        ]
      ]
    },
    "note": "First published: 2025-11-27; Recorded arXiv version: v2; Recorded version updated: 2026-03-04T12:15:06Z.",
    "title": "Training High-Level Schedulers with Execution-Feedback Reinforcement Learning for Long-Horizon GUI Automation",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2512.09396v1",
    "archive": "arXiv",
    "archive_location": "2512.09396",
    "author": [
      {
        "literal": "Zishu Wei"
      },
      {
        "literal": "Qixiang Ma"
      },
      {
        "literal": "Xavier Hu"
      },
      {
        "literal": "Yuhang Liu"
      },
      {
        "literal": "Hui Zang"
      },
      {
        "literal": "Yudong Zhao"
      },
      {
        "literal": "Tao Wang"
      },
      {
        "literal": "Shengyu Zhang"
      },
      {
        "literal": "Fei Wu"
      }
    ],
    "citation-key": "arxiv_2512_09396",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "49e76d19d0477bcc56f19e8170fed7e9cc80f9c4bafb5f03af3a2d172cb113bc",
        "submitted_at": "2025-12-10T07:40:23Z",
        "updated_at": "2025-12-10T07:40:23Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2512.09396v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2512.09396",
      "catalog_id": "arxiv:2512.09396",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-12-10",
      "identifiers": {
        "arxiv": "2512.09396"
      },
      "record_updated_at": "2025-12-10"
    },
    "genre": "preprint",
    "id": "arxiv:2512.09396",
    "issued": {
      "date-parts": [
        [
          2025,
          12,
          10
        ]
      ]
    },
    "note": "First published: 2025-12-10; Recorded arXiv version: v1; Recorded version updated: 2025-12-10T07:40:23Z.",
    "title": "GAIR: GUI Automation via Information-Joint Reasoning and Group Reflection",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2512.14014v2",
    "archive": "arXiv",
    "archive_location": "2512.14014",
    "author": [
      {
        "literal": "Shufan Li"
      },
      {
        "literal": "Konstantinos Kallidromitis"
      },
      {
        "literal": "Akash Gokul"
      },
      {
        "literal": "Yusuke Kato"
      },
      {
        "literal": "Kazuki Kozuka"
      },
      {
        "literal": "Aditya Grover"
      }
    ],
    "citation-key": "arxiv_2512_14014",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "e6228822cf64c6aeb1836bee52a6515123bd0d44d77cceb23e785bb4ff9aef47",
        "submitted_at": "2025-12-16T02:16:42Z",
        "updated_at": "2026-10-07T23:06:49Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2512.14014v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2512.14014",
      "catalog_id": "arxiv:2512.14014",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-12-16",
      "identifiers": {
        "arxiv": "2512.14014"
      },
      "record_updated_at": "2026-10-07"
    },
    "genre": "preprint",
    "id": "arxiv:2512.14014",
    "issued": {
      "date-parts": [
        [
          2025,
          12,
          16
        ]
      ]
    },
    "note": "First published: 2025-12-16; Recorded arXiv version: v2; Recorded version updated: 2026-10-07T23:06:49Z.",
    "title": "MobileWorldBench: Towards Semantic World Modeling For Mobile Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2512.23647v1",
    "archive": "arXiv",
    "archive_location": "2512.23647",
    "author": [
      {
        "literal": "Baixuan Li"
      },
      {
        "literal": "Jialong Wu"
      },
      {
        "literal": "Wenbiao Yin"
      },
      {
        "literal": "Kuan Li"
      },
      {
        "literal": "Zhongwang Zhang"
      },
      {
        "literal": "Huifeng Yin"
      },
      {
        "literal": "Zhengwei Tao"
      },
      {
        "literal": "Liwen Zhang"
      },
      {
        "literal": "Pengjun Xie"
      },
      {
        "literal": "Jingren Zhou"
      },
      {
        "literal": "Yong Jiang"
      }
    ],
    "citation-key": "arxiv_2512_23647",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "fde71dd935acb953bc5d01bcd48f0d6c3351db6a66126761baf134c114e7893c",
        "submitted_at": "2025-12-29T17:59:14Z",
        "updated_at": "2025-12-29T17:59:14Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2512.23647v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2512.23647",
      "catalog_id": "arxiv:2512.23647",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2025-12-29",
      "identifiers": {
        "arxiv": "2512.23647"
      },
      "record_updated_at": "2025-12-29"
    },
    "genre": "preprint",
    "id": "arxiv:2512.23647",
    "issued": {
      "date-parts": [
        [
          2025,
          12,
          29
        ]
      ]
    },
    "note": "First published: 2025-12-29; Recorded arXiv version: v1; Recorded version updated: 2025-12-29T17:59:14Z.",
    "title": "Nested Browser-Use Learning for Agentic Information Seeking",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2601.04035v2",
    "archive": "arXiv",
    "archive_location": "2601.04035",
    "author": [
      {
        "literal": "Yilin Cao"
      },
      {
        "literal": "Yufeng Zhong"
      },
      {
        "literal": "Zhixiong Zeng"
      },
      {
        "literal": "Siran Dai"
      },
      {
        "literal": "Liming Zheng"
      },
      {
        "literal": "Jing Huang"
      },
      {
        "literal": "Haibo Qiu"
      },
      {
        "literal": "Peng Shi"
      },
      {
        "literal": "Wenji Mao"
      }
    ],
    "citation-key": "arxiv_2601_04035",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "c40e3db3f538f089c19d7c8233ea01174d0591c3211a7acc412ee4df28e902a8",
        "submitted_at": "2026-01-07T15:51:44Z",
        "updated_at": "2026-08-31T13:27:52Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2601.04035v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2601.04035",
      "catalog_id": "arxiv:2601.04035",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-01-07",
      "identifiers": {
        "arxiv": "2601.04035"
      },
      "record_updated_at": "2026-08-31"
    },
    "genre": "preprint",
    "id": "arxiv:2601.04035",
    "issued": {
      "date-parts": [
        [
          2026,
          1,
          7
        ]
      ]
    },
    "note": "First published: 2026-01-07; Recorded arXiv version: v2; Recorded version updated: 2026-08-31T13:27:52Z.",
    "title": "MobileDreamer: Generative Sketch World Model for GUI Agent",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2601.07779v1",
    "archive": "arXiv",
    "archive_location": "2601.07779",
    "author": [
      {
        "literal": "Bowen Yang"
      },
      {
        "literal": "Kaiming Jin"
      },
      {
        "literal": "Zhenyu Wu"
      },
      {
        "literal": "Zhaoyang Liu"
      },
      {
        "literal": "Qiushi Sun"
      },
      {
        "literal": "Zehao Li"
      },
      {
        "literal": "JingJing Xie"
      },
      {
        "literal": "Zhoumianze Liu"
      },
      {
        "literal": "Fangzhi Xu"
      },
      {
        "literal": "Kanzhi Cheng"
      },
      {
        "literal": "Qingyun Li"
      },
      {
        "literal": "Yian Wang"
      },
      {
        "literal": "Yu Qiao"
      },
      {
        "literal": "Zun Wang"
      },
      {
        "literal": "Zichen Ding"
      }
    ],
    "citation-key": "arxiv_2601_07779",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "cfdaabaf3c6df80eb718e51dc666c54fb2ecf38ec520f6a71ccf5e5ce54248c8",
        "submitted_at": "2026-01-12T17:55:51Z",
        "updated_at": "2026-01-12T17:55:51Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2601.07779v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2601.07779",
      "catalog_id": "arxiv:2601.07779",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-01-12",
      "identifiers": {
        "arxiv": "2601.07779"
      },
      "record_updated_at": "2026-01-12"
    },
    "genre": "preprint",
    "id": "arxiv:2601.07779",
    "issued": {
      "date-parts": [
        [
          2026,
          1,
          12
        ]
      ]
    },
    "note": "First published: 2026-01-12; Recorded arXiv version: v1; Recorded version updated: 2026-01-12T17:55:51Z.",
    "title": "OS-Symphony: A Holistic Framework for Robust and Generalist Computer-Using Agent",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2601.08605v2",
    "archive": "arXiv",
    "archive_location": "2601.08605",
    "author": [
      {
        "literal": "Wenyuan Zhang"
      },
      {
        "literal": "Xinghua Zhang"
      },
      {
        "literal": "Haiyang Yu"
      },
      {
        "literal": "Shuaiyi Nie"
      },
      {
        "literal": "Bingli Wu"
      },
      {
        "literal": "Juwei Yue"
      },
      {
        "literal": "Tingwen Liu"
      },
      {
        "literal": "Yongbin Li"
      }
    ],
    "citation-key": "arxiv_2601_08605",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "73df8c5b417c3387530b7730be3aed8b69374c06fdd3573f56147b19359c6912",
        "submitted_at": "2026-01-13T14:48:34Z",
        "updated_at": "2026-04-15T14:59:17Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2601.08605v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2601.08605",
      "catalog_id": "arxiv:2601.08605",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-01-13",
      "identifiers": {
        "arxiv": "2601.08605"
      },
      "record_updated_at": "2026-04-15"
    },
    "genre": "preprint",
    "id": "arxiv:2601.08605",
    "issued": {
      "date-parts": [
        [
          2026,
          1,
          13
        ]
      ]
    },
    "note": "First published: 2026-01-13; Recorded arXiv version: v2; Recorded version updated: 2026-04-15T14:59:17Z.",
    "title": "ExpSeek: Self-Triggered Experience Seeking for Web Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2601.09923v3",
    "archive": "arXiv",
    "archive_location": "2601.09923",
    "author": [
      {
        "literal": "Hanna Foerster"
      },
      {
        "literal": "Tom Blanchard"
      },
      {
        "literal": "Kristina Nikolić"
      },
      {
        "literal": "Ilia Shumailov"
      },
      {
        "literal": "Cheng Zhang"
      },
      {
        "literal": "Robert Mullins"
      },
      {
        "literal": "Nicolas Papernot"
      },
      {
        "literal": "Florian Tramèr"
      },
      {
        "literal": "Yiren Zhao"
      }
    ],
    "citation-key": "arxiv_2601_09923",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "e32917bc92b65701eea06c09a2c681c7504418f8ce888af2b72012a1da50c9f6",
        "submitted_at": "2026-01-14T23:06:35Z",
        "updated_at": "2026-06-04T15:38:02Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2601.09923v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2601.09923",
      "catalog_id": "arxiv:2601.09923",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-01-14",
      "identifiers": {
        "arxiv": "2601.09923"
      },
      "record_updated_at": "2026-06-04"
    },
    "genre": "preprint",
    "id": "arxiv:2601.09923",
    "issued": {
      "date-parts": [
        [
          2026,
          1,
          14
        ]
      ]
    },
    "note": "First published: 2026-01-14; Recorded arXiv version: v3; Recorded version updated: 2026-06-04T15:38:02Z.",
    "title": "CaMeLs Can Use Computers Too: System-level Security for Computer Use Agents",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2601.17418v1",
    "archive": "arXiv",
    "archive_location": "2601.17418",
    "author": [
      {
        "literal": "Mingxian Yu"
      },
      {
        "literal": "Siqi Luo"
      },
      {
        "literal": "Xu Chen"
      }
    ],
    "citation-key": "arxiv_2601_17418",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "4ce0909d1c9081f099ae0dc0b0182e0442fd71fd58c62b13d486d154afc2187e",
        "submitted_at": "2026-01-24T11:35:37Z",
        "updated_at": "2026-01-24T11:35:37Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2601.17418v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2601.17418",
      "catalog_id": "arxiv:2601.17418",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-01-24",
      "identifiers": {
        "arxiv": "2601.17418"
      },
      "record_updated_at": "2026-01-24"
    },
    "genre": "preprint",
    "id": "arxiv:2601.17418",
    "issued": {
      "date-parts": [
        [
          2026,
          1,
          24
        ]
      ]
    },
    "note": "First published: 2026-01-24; Recorded arXiv version: v1; Recorded version updated: 2026-01-24T11:35:37Z.",
    "title": "GraphPilot: GUI Task Automation with One-Step LLM Reasoning Powered by Knowledge Graph",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2601.19306v1",
    "archive": "arXiv",
    "archive_location": "2601.19306",
    "author": [
      {
        "literal": "Sijia Li"
      },
      {
        "literal": "Xiaoyu Tan"
      },
      {
        "literal": "Shahir Ali"
      },
      {
        "literal": "Niels Schmidt"
      },
      {
        "literal": "Gengchen Ma"
      },
      {
        "literal": "Xihe Qiu"
      }
    ],
    "citation-key": "arxiv_2601_19306",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "223b9a915e5ea9849644318e1e5426e64ad8655ab46d0b007c30e1967130dc50",
        "submitted_at": "2026-01-27T07:46:05Z",
        "updated_at": "2026-01-27T07:46:05Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2601.19306v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2601.19306",
      "catalog_id": "arxiv:2601.19306",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-01-27",
      "identifiers": {
        "arxiv": "2601.19306"
      },
      "record_updated_at": "2026-01-27"
    },
    "genre": "preprint",
    "id": "arxiv:2601.19306",
    "issued": {
      "date-parts": [
        [
          2026,
          1,
          27
        ]
      ]
    },
    "note": "First published: 2026-01-27; Recorded arXiv version: v1; Recorded version updated: 2026-01-27T07:46:05Z.",
    "title": "Curiosity Driven Knowledge Retrieval for Mobile Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2601.21123v2",
    "archive": "arXiv",
    "archive_location": "2601.21123",
    "author": [
      {
        "literal": "Tianyi Chen"
      },
      {
        "literal": "Yinheng Li"
      },
      {
        "literal": "Michael Solodko"
      },
      {
        "literal": "Sen Wang"
      },
      {
        "literal": "Nan Jiang"
      },
      {
        "literal": "Tingyuan Cui"
      },
      {
        "literal": "Junheng Hao"
      },
      {
        "literal": "Jongwoo Ko"
      },
      {
        "literal": "Sara Abdali"
      },
      {
        "literal": "Leon Xu"
      },
      {
        "literal": "Suzhen Zheng"
      },
      {
        "literal": "Hao Fan"
      },
      {
        "literal": "Pashmina Cameron"
      },
      {
        "literal": "Justin Wagle"
      },
      {
        "literal": "Kazuhito Koishida"
      }
    ],
    "citation-key": "arxiv_2601_21123",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "5a3efa3cd0add437e327ea49896a3ee794daa178664dfb4096aba66fed27554b",
        "submitted_at": "2026-01-28T23:38:25Z",
        "updated_at": "2026-02-02T23:11:55Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2601.21123v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2601.21123",
      "catalog_id": "arxiv:2601.21123",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-01-28",
      "identifiers": {
        "arxiv": "2601.21123"
      },
      "record_updated_at": "2026-02-02"
    },
    "genre": "preprint",
    "id": "arxiv:2601.21123",
    "issued": {
      "date-parts": [
        [
          2026,
          1,
          28
        ]
      ]
    },
    "note": "First published: 2026-01-28; Recorded arXiv version: v2; Recorded version updated: 2026-02-02T23:11:55Z.",
    "title": "CUA-Skill: Develop Skills for Computer Using Agent",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2601.21352v1",
    "archive": "arXiv",
    "archive_location": "2601.21352",
    "author": [
      {
        "literal": "Ziyu Lu"
      },
      {
        "literal": "Tengjin Weng"
      },
      {
        "literal": "Yiying Yang"
      },
      {
        "literal": "Yuhang Zhao"
      },
      {
        "literal": "Xinxin Huang"
      },
      {
        "literal": "Wenhao Jiang"
      }
    ],
    "citation-key": "arxiv_2601_21352",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "fa36a9fbd0b189717b2a5311f06b6127d54f2eca09961cedd2a0b9d2a78b1b8b",
        "submitted_at": "2026-01-29T07:22:50Z",
        "updated_at": "2026-01-29T07:22:50Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2601.21352v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2601.21352",
      "catalog_id": "arxiv:2601.21352",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-01-29",
      "identifiers": {
        "arxiv": "2601.21352"
      },
      "record_updated_at": "2026-01-29"
    },
    "genre": "preprint",
    "id": "arxiv:2601.21352",
    "issued": {
      "date-parts": [
        [
          2026,
          1,
          29
        ]
      ]
    },
    "note": "First published: 2026-01-29; Recorded arXiv version: v1; Recorded version updated: 2026-01-29T07:22:50Z.",
    "title": "BEAP-Agent: Backtrackable Execution and Adaptive Planning for GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2602.01725v1",
    "archive": "arXiv",
    "archive_location": "2602.01725",
    "author": [
      {
        "literal": "Yurun Chen"
      },
      {
        "literal": "Zeyi Liao"
      },
      {
        "literal": "Ping Yin"
      },
      {
        "literal": "Taotao Xie"
      },
      {
        "literal": "Keting Yin"
      },
      {
        "literal": "Shengyu Zhang"
      }
    ],
    "citation-key": "arxiv_2602_01725",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d1a6f98c0ab0a221c169796d9680e4b2a7bf7a0775c0493254f47d5e0a9b3c4e",
        "submitted_at": "2026-02-02T07:04:06Z",
        "updated_at": "2026-02-02T07:04:06Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2602.01725v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.01725",
      "catalog_id": "arxiv:2602.01725",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-02",
      "identifiers": {
        "arxiv": "2602.01725"
      },
      "record_updated_at": "2026-02-02"
    },
    "genre": "preprint",
    "id": "arxiv:2602.01725",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          2
        ]
      ]
    },
    "note": "First published: 2026-02-02; Recorded arXiv version: v1; Recorded version updated: 2026-02-02T07:04:06Z.",
    "title": "SafePred: A Predictive Guardrail for Computer-Using Agents via World Models",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2602.02468v1",
    "archive": "arXiv",
    "archive_location": "2602.02468",
    "author": [
      {
        "literal": "Aiden Yiliu Li"
      },
      {
        "literal": "Xinyue Hao"
      },
      {
        "literal": "Shilong Liu"
      },
      {
        "literal": "Mengdi Wang"
      }
    ],
    "citation-key": "arxiv_2602_02468",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "a5b730fa81936d6a1b2b0a8b74a522a22de39b977929165c44a5050e490c301f",
        "submitted_at": "2026-02-02T18:50:07Z",
        "updated_at": "2026-02-02T18:50:07Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2602.02468v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.02468",
      "catalog_id": "arxiv:2602.02468",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-02",
      "identifiers": {
        "arxiv": "2602.02468"
      },
      "record_updated_at": "2026-02-02"
    },
    "genre": "preprint",
    "id": "arxiv:2602.02468",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          2
        ]
      ]
    },
    "note": "First published: 2026-02-02; Recorded arXiv version: v1; Recorded version updated: 2026-02-02T18:50:07Z.",
    "title": "Avenir-Web: Human-Experience-Imitating Multimodal Web Agents with Mixture of Grounding Experts",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2602.02995v1",
    "archive": "arXiv",
    "archive_location": "2602.02995",
    "author": [
      {
        "literal": "Sizhe Tang"
      },
      {
        "literal": "Rongqian Chen"
      },
      {
        "literal": "Tian Lan"
      }
    ],
    "citation-key": "arxiv_2602_02995",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "0a01499ce704bb7916a3157ff2d3fc755a1115d9b27c7e529c153bce360a5404",
        "submitted_at": "2026-02-03T02:07:12Z",
        "updated_at": "2026-02-03T02:07:12Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2602.02995v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.02995",
      "catalog_id": "arxiv:2602.02995",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-03",
      "identifiers": {
        "arxiv": "2602.02995"
      },
      "record_updated_at": "2026-02-03"
    },
    "genre": "preprint",
    "id": "arxiv:2602.02995",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          3
        ]
      ]
    },
    "note": "First published: 2026-02-03; Recorded arXiv version: v1; Recorded version updated: 2026-02-03T02:07:12Z.",
    "title": "Agent Alpha: Tree Search Unifying Generation, Exploration and Evaluation for Computer-Use Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2602.05354v3",
    "archive": "arXiv",
    "archive_location": "2602.05354",
    "author": [
      {
        "literal": "Syed Nazmus Sakib"
      },
      {
        "literal": "Nafiul Haque"
      },
      {
        "literal": "Tapodhir Karmakar Taton"
      },
      {
        "literal": "Shahrear Bin Amin"
      },
      {
        "literal": "Shifat E. Arman"
      }
    ],
    "citation-key": "arxiv_2602_05354",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "76fbeae086fb9ef21f820b02777a2ceb66ee37f8f1c978c7426589ea4e888e49",
        "submitted_at": "2026-02-05T06:24:23Z",
        "updated_at": "2026-09-08T14:40:01Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2602.05354v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.05354",
      "catalog_id": "arxiv:2602.05354",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-05",
      "identifiers": {
        "arxiv": "2602.05354"
      },
      "record_updated_at": "2026-09-08"
    },
    "genre": "preprint",
    "id": "arxiv:2602.05354",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          5
        ]
      ]
    },
    "note": "First published: 2026-02-05; Recorded arXiv version: v3; Recorded version updated: 2026-09-08T14:40:01Z.",
    "title": "Do Web Agents Investigate Before They Decide?",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2602.05429v1",
    "archive": "arXiv",
    "archive_location": "2602.05429",
    "author": [
      {
        "literal": "Rui Lv"
      },
      {
        "literal": "Juncheng Mo"
      },
      {
        "literal": "Tianyi Chu"
      },
      {
        "literal": "Chen Rao"
      },
      {
        "literal": "Hongyi Jing"
      },
      {
        "literal": "Jiajie Teng"
      },
      {
        "literal": "Jiafu Chen"
      },
      {
        "literal": "Shiqi Zhang"
      },
      {
        "literal": "Liangzi Ding"
      },
      {
        "literal": "Shuo Fang"
      },
      {
        "literal": "Huaizhong Lin"
      },
      {
        "literal": "Ziqiang Dang"
      },
      {
        "literal": "Chenguang Ma"
      },
      {
        "literal": "Lei Zhao"
      }
    ],
    "citation-key": "arxiv_2602_05429",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "84d822655c810958cff6035584542d7e151d765734ec7030f841c15561ee5677",
        "submitted_at": "2026-02-05T08:19:39Z",
        "updated_at": "2026-02-05T08:19:39Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2602.05429v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.05429",
      "catalog_id": "arxiv:2602.05429",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-05",
      "identifiers": {
        "arxiv": "2602.05429"
      },
      "record_updated_at": "2026-02-05"
    },
    "genre": "preprint",
    "id": "arxiv:2602.05429",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          5
        ]
      ]
    },
    "note": "First published: 2026-02-05; Recorded arXiv version: v1; Recorded version updated: 2026-02-05T08:19:39Z.",
    "title": "M$^2$-Miner: Multi-Agent Enhanced MCTS for Mobile GUI Agent Data Mining",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2602.07153v2",
    "archive": "arXiv",
    "archive_location": "2602.07153",
    "author": [
      {
        "literal": "Jinbiao Wei"
      },
      {
        "literal": "Yilun Zhao"
      },
      {
        "literal": "Kangqi Ni"
      },
      {
        "literal": "Arman Cohan"
      }
    ],
    "citation-key": "arxiv_2602_07153",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "ce896d4d846bf07fedc68b7f952532c4fefe24dcb49b51556e0d6c2bc648bb91",
        "submitted_at": "2026-02-06T19:55:26Z",
        "updated_at": "2026-04-12T06:09:23Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2602.07153v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.07153",
      "catalog_id": "arxiv:2602.07153",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-06",
      "identifiers": {
        "arxiv": "2602.07153"
      },
      "record_updated_at": "2026-04-12"
    },
    "genre": "preprint",
    "id": "arxiv:2602.07153",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          6
        ]
      ]
    },
    "note": "First published: 2026-02-06; Recorded arXiv version: v2; Recorded version updated: 2026-04-12T06:09:23Z.",
    "title": "ANCHOR: Branch-Point Data Generation for GUI Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2602.09662v1",
    "archive": "arXiv",
    "archive_location": "2602.09662",
    "author": [
      {
        "literal": "Deyang Jiang"
      },
      {
        "literal": "Jing Huang"
      },
      {
        "literal": "Xuanle Zhao"
      },
      {
        "literal": "Lei Chen"
      },
      {
        "literal": "Liming Zheng"
      },
      {
        "literal": "Fanfan Liu"
      },
      {
        "literal": "Haibo Qiu"
      },
      {
        "literal": "Peng Shi"
      },
      {
        "literal": "Zhixiong Zeng"
      }
    ],
    "citation-key": "arxiv_2602_09662",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "b5080050b85e38c8c03078c96c206e92d4ed63d106d55983563d51fcbc9f66c5",
        "submitted_at": "2026-02-10T11:16:57Z",
        "updated_at": "2026-02-10T11:16:57Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2602.09662v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.09662",
      "catalog_id": "arxiv:2602.09662",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-10",
      "identifiers": {
        "arxiv": "2602.09662"
      },
      "record_updated_at": "2026-02-10"
    },
    "genre": "preprint",
    "id": "arxiv:2602.09662",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          10
        ]
      ]
    },
    "note": "First published: 2026-02-10; Recorded arXiv version: v1; Recorded version updated: 2026-02-10T11:16:57Z.",
    "title": "TreeCUA: Efficiently Scaling GUI Automation with Tree-Structured Verifiable Evolution",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2602.11724v3",
    "archive": "arXiv",
    "archive_location": "2602.11724",
    "author": [
      {
        "literal": "Xiwen Teoh"
      },
      {
        "literal": "Yun Lin"
      },
      {
        "literal": "Duc-Minh Nguyen"
      },
      {
        "literal": "Ruofei Ren"
      },
      {
        "literal": "Wenjie Zhang"
      },
      {
        "literal": "Jin Song Dong"
      }
    ],
    "citation-key": "arxiv_2602_11724",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "2076fc8b0b7b37bdf43fe2e0b0cc5fa4c3a8cb7b44a41fdf80f59a8c05c7e1b3",
        "submitted_at": "2026-02-12T08:51:07Z",
        "updated_at": "2026-04-23T15:00:58Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2602.11724v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.11724",
      "catalog_id": "arxiv:2602.11724",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-12",
      "identifiers": {
        "arxiv": "2602.11724"
      },
      "record_updated_at": "2026-04-23"
    },
    "genre": "preprint",
    "id": "arxiv:2602.11724",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          12
        ]
      ]
    },
    "note": "First published: 2026-02-12; Recorded arXiv version: v3; Recorded version updated: 2026-04-23T15:00:58Z.",
    "title": "WebTestPilot: Agentic End-to-End Web Testing against Natural Language Specification by Inferring Oracles with Symbolized GUI Elements",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2602.12276v2",
    "archive": "arXiv",
    "archive_location": "2602.12276",
    "author": [
      {
        "literal": "Nicholas Lee"
      },
      {
        "literal": "Lutfi Eren Erdogan"
      },
      {
        "literal": "Chris Joseph John"
      },
      {
        "literal": "Surya Krishnapillai"
      },
      {
        "literal": "Michael W. Mahoney"
      },
      {
        "literal": "Kurt Keutzer"
      },
      {
        "literal": "Amir Gholami"
      }
    ],
    "citation-key": "arxiv_2602_12276",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "c64cd65fc7fec601658baf7836bdbe9c0e494bde2b0a4b66ec9d3090ee106b1f",
        "submitted_at": "2026-02-12T18:58:30Z",
        "updated_at": "2026-08-14T21:25:10Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2602.12276v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.12276",
      "catalog_id": "arxiv:2602.12276",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-12",
      "identifiers": {
        "arxiv": "2602.12276"
      },
      "record_updated_at": "2026-08-14"
    },
    "genre": "preprint",
    "id": "arxiv:2602.12276",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          12
        ]
      ]
    },
    "note": "First published: 2026-02-12; Recorded arXiv version: v2; Recorded version updated: 2026-08-14T21:25:10Z.",
    "title": "Agentic Test-Time Scaling for WebAgents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2602.12430v4",
    "archive": "arXiv",
    "archive_location": "2602.12430",
    "author": [
      {
        "literal": "Renjun Xu"
      },
      {
        "literal": "Yang Yan"
      }
    ],
    "citation-key": "arxiv_2602_12430",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "f0edc70265a8d46a7c22a9c6384de8aa19f79dbb6987f8c23f4a260647a6a3f0",
        "submitted_at": "2026-02-12T21:33:25Z",
        "updated_at": "2026-06-02T16:14:54Z",
        "version": 4,
        "versioned_url": "https://arxiv.org/abs/2602.12430v4"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.12430",
      "catalog_id": "arxiv:2602.12430",
      "catalog_kind": "survey",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2026-02-12",
      "identifiers": {
        "arxiv": "2602.12430"
      },
      "record_updated_at": "2026-06-02"
    },
    "genre": "preprint",
    "id": "arxiv:2602.12430",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          12
        ]
      ]
    },
    "note": "First published: 2026-02-12; Recorded arXiv version: v4; Recorded version updated: 2026-06-02T16:14:54Z.",
    "title": "Agent Skills for Large Language Models: Architecture, Acquisition, Security, and the Path Forward",
    "type": "article",
    "version": "4"
  },
  {
    "URL": "https://arxiv.org/abs/2602.13559v2",
    "archive": "arXiv",
    "archive_location": "2602.13559",
    "author": [
      {
        "literal": "Yuyu Guo"
      },
      {
        "literal": "Wenjie Yang"
      },
      {
        "literal": "Siyuan Yang"
      },
      {
        "literal": "Ziyang Liu"
      },
      {
        "literal": "Cheng Chen"
      },
      {
        "literal": "Yuan Wei"
      },
      {
        "literal": "Yun Hu"
      },
      {
        "literal": "Yang Huang"
      },
      {
        "literal": "Guoliang Hao"
      },
      {
        "literal": "Dongsheng Yuan"
      },
      {
        "literal": "Jianming Wang"
      },
      {
        "literal": "Xin Chen"
      },
      {
        "literal": "Hang Yu"
      },
      {
        "literal": "Lei Lei"
      },
      {
        "literal": "Peng Di"
      }
    ],
    "citation-key": "arxiv_2602_13559",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "04298ed8a4bc0f2811f52eeee6da0a13faa265bb3c5cbc5ea6be573a77c21f27",
        "submitted_at": "2026-02-14T02:33:55Z",
        "updated_at": "2026-04-30T02:05:10Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2602.13559v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.13559",
      "catalog_id": "arxiv:2602.13559",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-14",
      "identifiers": {
        "arxiv": "2602.13559"
      },
      "record_updated_at": "2026-04-30"
    },
    "genre": "preprint",
    "id": "arxiv:2602.13559",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          14
        ]
      ]
    },
    "note": "First published: 2026-02-14; Recorded arXiv version: v2; Recorded version updated: 2026-04-30T02:05:10Z.",
    "title": "OpAgent: Operator Agent for Web Navigation",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2602.16313v2",
    "archive": "arXiv",
    "archive_location": "2602.16313",
    "author": [
      {
        "literal": "Zexue He"
      },
      {
        "literal": "Yu Wang"
      },
      {
        "literal": "Churan Zhi"
      },
      {
        "literal": "Yuanzhe Hu"
      },
      {
        "literal": "Tzu-Ping Chen"
      },
      {
        "literal": "Lang Yin"
      },
      {
        "literal": "Ze Chen"
      },
      {
        "literal": "Tong Arthur Wu"
      },
      {
        "literal": "Siru Ouyang"
      },
      {
        "literal": "Zihan Wang"
      },
      {
        "literal": "Jiaxin Pei"
      },
      {
        "literal": "Julian McAuley"
      },
      {
        "literal": "Yejin Choi"
      },
      {
        "literal": "Alex Pentland"
      }
    ],
    "citation-key": "arxiv_2602_16313",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "2317a94d54b4458a1b7c5e92754aeaefcbf270aabfa0bd8ea333c673cd4596b3",
        "submitted_at": "2026-02-18T09:49:14Z",
        "updated_at": "2026-09-17T22:55:03Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2602.16313v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.16313",
      "catalog_id": "arxiv:2602.16313",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-18",
      "identifiers": {
        "arxiv": "2602.16313"
      },
      "record_updated_at": "2026-09-17"
    },
    "genre": "preprint",
    "id": "arxiv:2602.16313",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          18
        ]
      ]
    },
    "note": "First published: 2026-02-18; Recorded arXiv version: v2; Recorded version updated: 2026-09-17T22:55:03Z.",
    "title": "MemoryArena: Benchmarking Agent Memory in Interdependent Multi-Session Agentic Tasks",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2602.17365v1",
    "archive": "arXiv",
    "archive_location": "2602.17365",
    "author": [
      {
        "literal": "Yiming Guan"
      },
      {
        "literal": "Rui Yu"
      },
      {
        "literal": "John Zhang"
      },
      {
        "literal": "Lu Wang"
      },
      {
        "literal": "Chaoyun Zhang"
      },
      {
        "literal": "Liqun Li"
      },
      {
        "literal": "Bo Qiao"
      },
      {
        "literal": "Si Qin"
      },
      {
        "literal": "He Huang"
      },
      {
        "literal": "Fangkai Yang"
      },
      {
        "literal": "Pu Zhao"
      },
      {
        "literal": "Lukas Wutschitz"
      },
      {
        "literal": "Samuel Kessler"
      },
      {
        "literal": "Huseyin A Inan"
      },
      {
        "literal": "Robert Sim"
      },
      {
        "literal": "Saravan Rajmohan"
      },
      {
        "literal": "Qingwei Lin"
      },
      {
        "literal": "Dongmei Zhang"
      }
    ],
    "citation-key": "arxiv_2602_17365",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "3d2237dd43f4c876ba69b68c1871ca3efd3aa501e93c86655eb2de6d9f35d2e8",
        "submitted_at": "2026-02-19T13:48:29Z",
        "updated_at": "2026-02-19T13:48:29Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2602.17365v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.17365",
      "catalog_id": "arxiv:2602.17365",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-19",
      "identifiers": {
        "arxiv": "2602.17365"
      },
      "record_updated_at": "2026-02-19"
    },
    "genre": "preprint",
    "id": "arxiv:2602.17365",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          19
        ]
      ]
    },
    "note": "First published: 2026-02-19; Recorded arXiv version: v1; Recorded version updated: 2026-02-19T13:48:29Z.",
    "title": "Computer-Using World Model",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2602.20502v2",
    "archive": "arXiv",
    "archive_location": "2602.20502",
    "author": [
      {
        "literal": "Hongbin Zhong"
      },
      {
        "literal": "Fazle Faisal"
      },
      {
        "literal": "Luis França"
      },
      {
        "literal": "Tanakorn Leesatapornwongsa"
      },
      {
        "literal": "Adriana Szekeres"
      },
      {
        "literal": "Kexin Rong"
      },
      {
        "literal": "Suman Nath"
      }
    ],
    "citation-key": "arxiv_2602_20502",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "e179c833d43a88098d576e31e726eb3e4289f728659697364d0b448282df3fc1",
        "submitted_at": "2026-02-24T03:03:18Z",
        "updated_at": "2026-09-28T17:40:20Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2602.20502v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2602.20502",
      "catalog_id": "arxiv:2602.20502",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-02-24",
      "identifiers": {
        "arxiv": "2602.20502"
      },
      "record_updated_at": "2026-09-28"
    },
    "genre": "preprint",
    "id": "arxiv:2602.20502",
    "issued": {
      "date-parts": [
        [
          2026,
          2,
          24
        ]
      ]
    },
    "note": "First published: 2026-02-24; Recorded arXiv version: v2; Recorded version updated: 2026-09-28T17:40:20Z.",
    "title": "ActionEngine: From Reactive to Programmatic Web Agents via State Machine Memory",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2603.05295v3",
    "archive": "arXiv",
    "archive_location": "2603.05295",
    "author": [
      {
        "literal": "Sicheng Fan"
      },
      {
        "literal": "Rui Wan"
      },
      {
        "literal": "Yifei Leng"
      },
      {
        "literal": "Gaoning Liang"
      },
      {
        "literal": "Li Ling"
      },
      {
        "literal": "Yanyi Shang"
      },
      {
        "literal": "Dehan Kong"
      }
    ],
    "citation-key": "arxiv_2603_05295",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "db52246d9f5f305046e6dd14f2ef15440c371678e1db0d05282c84a3667c6f10",
        "submitted_at": "2026-03-05T15:37:34Z",
        "updated_at": "2026-04-14T08:41:56Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2603.05295v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2603.05295",
      "catalog_id": "arxiv:2603.05295",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-03-05",
      "identifiers": {
        "arxiv": "2603.05295"
      },
      "record_updated_at": "2026-04-14"
    },
    "genre": "preprint",
    "id": "arxiv:2603.05295",
    "issued": {
      "date-parts": [
        [
          2026,
          3,
          5
        ]
      ]
    },
    "note": "First published: 2026-03-05; Recorded arXiv version: v3; Recorded version updated: 2026-04-14T08:41:56Z.",
    "title": "WebChain: A Large-Scale Human-Annotated Dataset of Real-World Web Interaction Traces",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2603.07978v1",
    "archive": "arXiv",
    "archive_location": "2603.07978",
    "author": [
      {
        "literal": "Jiateng Liu"
      },
      {
        "literal": "Zhenhailong Wang"
      },
      {
        "literal": "Rushi Wang"
      },
      {
        "literal": "Bingxuan Li"
      },
      {
        "literal": "Jeonghwan Kim"
      },
      {
        "literal": "Aditi Tiwari"
      },
      {
        "literal": "Pengfei Yu"
      },
      {
        "literal": "Denghui Zhang"
      },
      {
        "literal": "Heng Ji"
      }
    ],
    "citation-key": "arxiv_2603_07978",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "77eb68f31d616b8e542dbb41ada67f28bdc21f29f7da3f04a0077c0fa2d10196",
        "submitted_at": "2026-03-09T05:27:56Z",
        "updated_at": "2026-03-09T05:27:56Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2603.07978v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2603.07978",
      "catalog_id": "arxiv:2603.07978",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-03-09",
      "identifiers": {
        "arxiv": "2603.07978"
      },
      "record_updated_at": "2026-03-09"
    },
    "genre": "preprint",
    "id": "arxiv:2603.07978",
    "issued": {
      "date-parts": [
        [
          2026,
          3,
          9
        ]
      ]
    },
    "note": "First published: 2026-03-09; Recorded arXiv version: v1; Recorded version updated: 2026-03-09T05:27:56Z.",
    "title": "OSExpert: Computer-Use Agents Learning Professional Skills via Exploration",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2603.10268v1",
    "archive": "arXiv",
    "archive_location": "2603.10268",
    "author": [
      {
        "literal": "Syed Yusuf Ahmed"
      },
      {
        "literal": "Shiwei Feng"
      },
      {
        "literal": "Chanwoo Bae"
      },
      {
        "literal": "Calix Barrus Xiangyu Zhang"
      }
    ],
    "citation-key": "arxiv_2603_10268",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "8ad18d7fe815f5529fef663130b7ce8841763589e7f4319c0c7556ec2df1da03",
        "submitted_at": "2026-03-10T22:56:03Z",
        "updated_at": "2026-03-10T22:56:03Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2603.10268v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2603.10268",
      "catalog_id": "arxiv:2603.10268",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-03-10",
      "identifiers": {
        "arxiv": "2603.10268"
      },
      "record_updated_at": "2026-03-10"
    },
    "genre": "preprint",
    "id": "arxiv:2603.10268",
    "issued": {
      "date-parts": [
        [
          2026,
          3,
          10
        ]
      ]
    },
    "note": "First published: 2026-03-10; Recorded arXiv version: v1; Recorded version updated: 2026-03-10T22:56:03Z.",
    "title": "SpecOps: A Fully Automated AI Agent Testing Framework in Real-World GUI Environments",
    "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/2603.20340v3",
    "archive": "arXiv",
    "archive_location": "2603.20340",
    "author": [
      {
        "literal": "Zijian Lu"
      },
      {
        "literal": "Yiping Zuo"
      },
      {
        "literal": "Yupeng Nie"
      },
      {
        "literal": "Xin He"
      },
      {
        "literal": "Weibei Fan"
      },
      {
        "literal": "Lianyong Qi"
      },
      {
        "literal": "Shi Jin"
      }
    ],
    "citation-key": "arxiv_2603_20340",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "8e8182e64c7a57a2dd6eb381762bc9a60b08e13ba5a337d0e9125ab6a5518b6c",
        "submitted_at": "2026-03-20T09:25:50Z",
        "updated_at": "2026-04-15T07:34:34Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2603.20340v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2603.20340",
      "catalog_id": "arxiv:2603.20340",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-03-20",
      "identifiers": {
        "arxiv": "2603.20340"
      },
      "record_updated_at": "2026-04-15"
    },
    "genre": "preprint",
    "id": "arxiv:2603.20340",
    "issued": {
      "date-parts": [
        [
          2026,
          3,
          20
        ]
      ]
    },
    "note": "First published: 2026-03-20; Recorded arXiv version: v3; Recorded version updated: 2026-04-15T07:34:34Z.",
    "title": "ContractSkill: Repairable Contract-Based Skills for Multimodal Web Agents",
    "type": "article",
    "version": "3"
  },
  {
    "URL": "https://arxiv.org/abs/2603.22529v1",
    "archive": "arXiv",
    "archive_location": "2603.22529",
    "author": [
      {
        "literal": "Shoubin Yu"
      },
      {
        "literal": "Lei Shu"
      },
      {
        "literal": "Antoine Yang"
      },
      {
        "literal": "Yao Fu"
      },
      {
        "literal": "Srinivas Sunkara"
      },
      {
        "literal": "Maria Wang"
      },
      {
        "literal": "Jindong Chen"
      },
      {
        "literal": "Mohit Bansal"
      },
      {
        "literal": "Boqing Gong"
      }
    ],
    "citation-key": "arxiv_2603_22529",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "3e4ed7391969359c2594e8b96b8a3a559ef98d2c84a10c1a2c39b3bd962b58c5",
        "submitted_at": "2026-03-23T19:47:58Z",
        "updated_at": "2026-03-23T19:47:58Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2603.22529v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2603.22529",
      "catalog_id": "arxiv:2603.22529",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-03-23",
      "identifiers": {
        "arxiv": "2603.22529"
      },
      "record_updated_at": "2026-03-23"
    },
    "genre": "preprint",
    "id": "arxiv:2603.22529",
    "issued": {
      "date-parts": [
        [
          2026,
          3,
          23
        ]
      ]
    },
    "note": "First published: 2026-03-23; Recorded arXiv version: v1; Recorded version updated: 2026-03-23T19:47:58Z.",
    "title": "Ego2Web: A Web Agent Benchmark Grounded in Egocentric Videos",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2603.24157v1",
    "archive": "arXiv",
    "archive_location": "2603.24157",
    "author": [
      {
        "literal": "Akash Ghosh"
      },
      {
        "literal": "Tajamul Ashraf"
      },
      {
        "literal": "Rishu Kumar Singh"
      },
      {
        "literal": "Numan Saeed"
      },
      {
        "literal": "Sriparna Saha"
      },
      {
        "literal": "Xiuying Chen"
      },
      {
        "literal": "Salman Khan"
      }
    ],
    "citation-key": "arxiv_2603_24157",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "8fb6a294616e1be46c5da05f60cda7d43c05d8a1ce71116c5e30cb94507cd56f",
        "submitted_at": "2026-03-25T10:25:48Z",
        "updated_at": "2026-03-25T10:25:48Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2603.24157v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2603.24157",
      "catalog_id": "arxiv:2603.24157",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-03-25",
      "identifiers": {
        "arxiv": "2603.24157"
      },
      "record_updated_at": "2026-03-25"
    },
    "genre": "preprint",
    "id": "arxiv:2603.24157",
    "issued": {
      "date-parts": [
        [
          2026,
          3,
          25
        ]
      ]
    },
    "note": "First published: 2026-03-25; Recorded arXiv version: v1; Recorded version updated: 2026-03-25T10:25:48Z.",
    "title": "CarePilot: A Multi-Agent Framework for Long-Horizon Computer Task Automation in Healthcare",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2603.26266v3",
    "archive": "arXiv",
    "archive_location": "2603.26266",
    "author": [
      {
        "literal": "Rui Xie"
      },
      {
        "literal": "Zhi Gao"
      },
      {
        "literal": "Chenrui Shi"
      },
      {
        "literal": "Zirui Shang"
      },
      {
        "literal": "Lu Chen"
      },
      {
        "literal": "Qing Li"
      }
    ],
    "citation-key": "arxiv_2603_26266",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "b388086aeb8ac47e386099e5cd68cb89bb03978ee7d5897950181035d0abb723",
        "submitted_at": "2026-03-27T10:33:08Z",
        "updated_at": "2026-06-30T04:47:42Z",
        "version": 3,
        "versioned_url": "https://arxiv.org/abs/2603.26266v3"
      },
      "canonical_source_url": "https://arxiv.org/abs/2603.26266",
      "catalog_id": "arxiv:2603.26266",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-03-27",
      "identifiers": {
        "arxiv": "2603.26266"
      },
      "record_updated_at": "2026-06-30"
    },
    "genre": "preprint",
    "id": "arxiv:2603.26266",
    "issued": {
      "date-parts": [
        [
          2026,
          3,
          27
        ]
      ]
    },
    "note": "First published: 2026-03-27; Recorded arXiv version: v3; Recorded version updated: 2026-06-30T04:47:42Z.",
    "title": "GUIDE: Resolving Domain Bias in GUI Agents through Real-Time Web Video Retrieval and Plug-and-Play Annotation",
    "type": "article",
    "version": "3"
  },
  {
    "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/2604.13318v2",
    "archive": "arXiv",
    "archive_location": "2604.13318",
    "author": [
      {
        "literal": "Zhaoyang Wang"
      },
      {
        "literal": "Qianhui Wu"
      },
      {
        "literal": "Xuchao Zhang"
      },
      {
        "literal": "Chaoyun Zhang"
      },
      {
        "literal": "Wenlin Yao"
      },
      {
        "literal": "Fazle Elahi Faisal"
      },
      {
        "literal": "Baolin Peng"
      },
      {
        "literal": "Si Qin"
      },
      {
        "literal": "Suman Nath"
      },
      {
        "literal": "Qingwei Lin"
      },
      {
        "literal": "Chetan Bansal"
      },
      {
        "literal": "Dongmei Zhang"
      },
      {
        "literal": "Saravan Rajmohan"
      },
      {
        "literal": "Jianfeng Gao"
      },
      {
        "literal": "Huaxiu Yao"
      }
    ],
    "citation-key": "arxiv_2604_13318",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "bf709a3d7096bc07218206edb1accfdda64036a39ea770e637b3e073dd9a76c8",
        "submitted_at": "2026-04-14T21:48:15Z",
        "updated_at": "2026-08-31T07:02:14Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2604.13318v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2604.13318",
      "catalog_id": "arxiv:2604.13318",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-04-14",
      "identifiers": {
        "arxiv": "2604.13318"
      },
      "record_updated_at": "2026-08-31"
    },
    "genre": "preprint",
    "id": "arxiv:2604.13318",
    "issued": {
      "date-parts": [
        [
          2026,
          4,
          14
        ]
      ]
    },
    "note": "First published: 2026-04-14; Recorded arXiv version: v2; Recorded version updated: 2026-08-31T07:02:14Z.",
    "title": "WebXSkill: Skill Learning for Autonomous Web Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2604.13822v1",
    "archive": "arXiv",
    "archive_location": "2604.13822",
    "author": [
      {
        "literal": "Zhengxi Lu"
      },
      {
        "literal": "Fei Tang"
      },
      {
        "literal": "Guangyi Liu"
      },
      {
        "literal": "Kaitao Song"
      },
      {
        "literal": "Xu Tan"
      },
      {
        "literal": "Jin Ma"
      },
      {
        "literal": "Wenqi Zhang"
      },
      {
        "literal": "Weiming Lu"
      },
      {
        "literal": "Jun Xiao"
      },
      {
        "literal": "Yueting Zhuang"
      },
      {
        "literal": "Yongliang Shen"
      }
    ],
    "citation-key": "arxiv_2604_13822",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d6e81b448c9c22e2fd25dce519db3fdc0585b0265b94d448751d7a37ed910dfa",
        "submitted_at": "2026-04-15T12:55:46Z",
        "updated_at": "2026-04-15T12:55:46Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2604.13822v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2604.13822",
      "catalog_id": "arxiv:2604.13822",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-04-15",
      "identifiers": {
        "arxiv": "2604.13822"
      },
      "record_updated_at": "2026-04-15"
    },
    "genre": "preprint",
    "id": "arxiv:2604.13822",
    "issued": {
      "date-parts": [
        [
          2026,
          4,
          15
        ]
      ]
    },
    "note": "First published: 2026-04-15; Recorded arXiv version: v1; Recorded version updated: 2026-04-15T12:55:46Z.",
    "title": "UI-Copilot: Advancing Long-Horizon GUI Automation via Tool-Integrated Policy Optimization",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2604.17821v2",
    "archive": "arXiv",
    "archive_location": "2604.17821",
    "author": [
      {
        "literal": "Lingfeng Zhang"
      },
      {
        "literal": "Yongan Sun"
      },
      {
        "literal": "Jinpeng Hu"
      },
      {
        "literal": "Hui Ma"
      },
      {
        "literal": "Yang Ying"
      },
      {
        "literal": "Kuien Liu"
      },
      {
        "literal": "Zenglin Shi"
      },
      {
        "literal": "Meng Wang"
      }
    ],
    "citation-key": "arxiv_2604_17821",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "123577fbb4431f22bc18d2d32187744268f9d63b3149e9e1775fb44673771feb",
        "submitted_at": "2026-04-20T05:19:49Z",
        "updated_at": "2026-04-21T03:15:01Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2604.17821v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2604.17821",
      "catalog_id": "arxiv:2604.17821",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-04-20",
      "identifiers": {
        "arxiv": "2604.17821"
      },
      "record_updated_at": "2026-04-21"
    },
    "genre": "preprint",
    "id": "arxiv:2604.17821",
    "issued": {
      "date-parts": [
        [
          2026,
          4,
          20
        ]
      ]
    },
    "note": "First published: 2026-04-20; Recorded arXiv version: v2; Recorded version updated: 2026-04-21T03:15:01Z.",
    "title": "WebUncertainty: Dual-Level Uncertainty Driven Planning and Reasoning For Autonomous Web Agent",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2605.12294v1",
    "archive": "arXiv",
    "archive_location": "2605.12294",
    "author": [
      {
        "literal": "Zerui Qin"
      },
      {
        "literal": "Sheng Yue"
      },
      {
        "literal": "Xingyuan Hua"
      },
      {
        "literal": "Yongjian Fu"
      },
      {
        "literal": "Ju Ren"
      }
    ],
    "citation-key": "arxiv_2605_12294",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "388eac2721929fc8211f40cc4edde61be0b6e6145fa60fbaa444ec80822fcc36",
        "submitted_at": "2026-05-12T15:48:44Z",
        "updated_at": "2026-05-12T15:48:44Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2605.12294v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2605.12294",
      "catalog_id": "arxiv:2605.12294",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-05-12",
      "identifiers": {
        "arxiv": "2605.12294"
      },
      "record_updated_at": "2026-05-12"
    },
    "genre": "preprint",
    "id": "arxiv:2605.12294",
    "issued": {
      "date-parts": [
        [
          2026,
          5,
          12
        ]
      ]
    },
    "note": "First published: 2026-05-12; Recorded arXiv version: v1; Recorded version updated: 2026-05-12T15:48:44Z.",
    "title": "Executable Agentic Memory for GUI Agent",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2605.12481v2",
    "archive": "arXiv",
    "archive_location": "2605.12481",
    "author": [
      {
        "literal": "Xuhao Hu"
      },
      {
        "literal": "Xi Zhang"
      },
      {
        "literal": "Haiyang Xu"
      },
      {
        "literal": "Kyle Qiao"
      },
      {
        "literal": "Jingyi Yang"
      },
      {
        "literal": "Xuanjing Huang"
      },
      {
        "literal": "Jing Shao"
      },
      {
        "literal": "Ming Yan"
      },
      {
        "literal": "Jieping Ye"
      }
    ],
    "citation-key": "arxiv_2605_12481",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "09128e90b925ac0cd2cf6e967853180bbef1e97a0960c22065f1b0b50c30d58b",
        "submitted_at": "2026-05-12T17:57:04Z",
        "updated_at": "2026-09-19T14:53:46Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2605.12481v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2605.12481",
      "catalog_id": "arxiv:2605.12481",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-05-12",
      "identifiers": {
        "arxiv": "2605.12481"
      },
      "record_updated_at": "2026-09-19"
    },
    "genre": "preprint",
    "id": "arxiv:2605.12481",
    "issued": {
      "date-parts": [
        [
          2026,
          5,
          12
        ]
      ]
    },
    "note": "First published: 2026-05-12; Recorded arXiv version: v2; Recorded version updated: 2026-09-19T14:53:46Z.",
    "title": "ToolCUA: Towards Optimal GUI-Tool Path Orchestration for Computer Use Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2605.14290v1",
    "archive": "arXiv",
    "archive_location": "2605.14290",
    "author": [
      {
        "literal": "Julien Piet"
      },
      {
        "literal": "Annabella Chow"
      },
      {
        "literal": "Yiwei Hou"
      },
      {
        "literal": "Muxi Lyu"
      },
      {
        "literal": "Sylvie Venuto"
      },
      {
        "literal": "Jinhao Zhu"
      },
      {
        "literal": "Raluca Ada Popa"
      },
      {
        "literal": "David Wagner"
      }
    ],
    "citation-key": "arxiv_2605_14290",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "572be49190554bbfb6accac4c87ef91608f88e90d06250f229db2922e19fcd56",
        "submitted_at": "2026-05-14T02:48:57Z",
        "updated_at": "2026-05-14T02:48:57Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2605.14290v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2605.14290",
      "catalog_id": "arxiv:2605.14290",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2026-05-14",
      "identifiers": {
        "arxiv": "2605.14290"
      },
      "record_updated_at": "2026-05-14"
    },
    "genre": "preprint",
    "id": "arxiv:2605.14290",
    "issued": {
      "date-parts": [
        [
          2026,
          5,
          14
        ]
      ]
    },
    "note": "First published: 2026-05-14; Recorded arXiv version: v1; Recorded version updated: 2026-05-14T02:48:57Z.",
    "title": "Web Agents Should Adopt the Plan-Then-Execute Paradigm",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2605.15542v1",
    "archive": "arXiv",
    "archive_location": "2605.15542",
    "author": [
      {
        "literal": "Yichao Liu"
      },
      {
        "literal": "Huawen Shen"
      },
      {
        "literal": "Liu Yu"
      },
      {
        "literal": "Shiyu Liu"
      },
      {
        "literal": "Zeyu Chen"
      },
      {
        "literal": "Yu Zhou"
      }
    ],
    "citation-key": "arxiv_2605_15542",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "fd85c0d0a8b6bc5c687bc301b1b84d6a019c5713012890e0a23c08ae2d83a1e6",
        "submitted_at": "2026-05-15T02:27:41Z",
        "updated_at": "2026-05-15T02:27:41Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2605.15542v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2605.15542",
      "catalog_id": "arxiv:2605.15542",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-05-15",
      "identifiers": {
        "arxiv": "2605.15542"
      },
      "record_updated_at": "2026-05-15"
    },
    "genre": "preprint",
    "id": "arxiv:2605.15542",
    "issued": {
      "date-parts": [
        [
          2026,
          5,
          15
        ]
      ]
    },
    "note": "First published: 2026-05-15; Recorded arXiv version: v1; Recorded version updated: 2026-05-15T02:27:41Z.",
    "title": "DRS-GUI: Dynamic Region Search for Training-Free GUI Grounding",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2605.18048v1",
    "archive": "arXiv",
    "archive_location": "2605.18048",
    "author": [
      {
        "literal": "Jingjing Liu"
      },
      {
        "literal": "Ziye Huang"
      },
      {
        "literal": "Zihao Cheng"
      },
      {
        "literal": "Zeming Liu"
      },
      {
        "literal": "Jiahong Wu"
      },
      {
        "literal": "Yuhang Guo"
      },
      {
        "literal": "Kehai Chen"
      },
      {
        "literal": "Yunhong Wang"
      },
      {
        "literal": "Haifeng Wang"
      }
    ],
    "citation-key": "arxiv_2605_18048",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "017909a697c6c661d3486d960042f4d691f70a660a60083e70512de34690cb51",
        "submitted_at": "2026-05-18T08:36:04Z",
        "updated_at": "2026-05-18T08:36:04Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2605.18048v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2605.18048",
      "catalog_id": "arxiv:2605.18048",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-05-18",
      "identifiers": {
        "arxiv": "2605.18048"
      },
      "record_updated_at": "2026-05-18"
    },
    "genre": "preprint",
    "id": "arxiv:2605.18048",
    "issued": {
      "date-parts": [
        [
          2026,
          5,
          18
        ]
      ]
    },
    "note": "First published: 2026-05-18; Recorded arXiv version: v1; Recorded version updated: 2026-05-18T08:36:04Z.",
    "title": "DocOS: Towards Proactive Document-Guided Actions in GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2605.21470v2",
    "archive": "arXiv",
    "archive_location": "2605.21470",
    "author": [
      {
        "literal": "Caleb Winston"
      },
      {
        "literal": "Ron Yifeng Wang"
      },
      {
        "literal": "Azalia Mirhoseini"
      },
      {
        "literal": "Christos Kozyrakis"
      }
    ],
    "citation-key": "arxiv_2605_21470",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "9b73bf7776e280a9040b1d7d1e29957913723253014c1dfe84f64916b2a08e5e",
        "submitted_at": "2026-05-20T17:54:27Z",
        "updated_at": "2026-05-29T17:54:05Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2605.21470v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2605.21470",
      "catalog_id": "arxiv:2605.21470",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-05-20",
      "identifiers": {
        "arxiv": "2605.21470"
      },
      "record_updated_at": "2026-05-29"
    },
    "genre": "preprint",
    "id": "arxiv:2605.21470",
    "issued": {
      "date-parts": [
        [
          2026,
          5,
          20
        ]
      ]
    },
    "note": "First published: 2026-05-20; Recorded arXiv version: v2; Recorded version updated: 2026-05-29T17:54:05Z.",
    "title": "Agent JIT Compilation for Latency-Optimizing Web Agent Planning and Scheduling",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2605.28534v1",
    "archive": "arXiv",
    "archive_location": "2605.28534",
    "author": [
      {
        "literal": "Zheng Wu"
      },
      {
        "literal": "Chengcheng Han"
      },
      {
        "literal": "Zhengxi Lu"
      },
      {
        "literal": "Tianjie Ju"
      },
      {
        "literal": "Yanyu Chen"
      },
      {
        "literal": "Qi Gu"
      },
      {
        "literal": "Xunliang Cai"
      },
      {
        "literal": "Zhuosheng Zhang"
      }
    ],
    "citation-key": "arxiv_2605_28534",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "78a5f5d7df9a8b6f1dbdb5693849c728a4b97e1f25c6e2e24d0c2418628c64e7",
        "submitted_at": "2026-05-27T14:29:41Z",
        "updated_at": "2026-05-27T14:29:41Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2605.28534v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2605.28534",
      "catalog_id": "arxiv:2605.28534",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-05-27",
      "identifiers": {
        "arxiv": "2605.28534"
      },
      "record_updated_at": "2026-05-27"
    },
    "genre": "preprint",
    "id": "arxiv:2605.28534",
    "issued": {
      "date-parts": [
        [
          2026,
          5,
          27
        ]
      ]
    },
    "note": "First published: 2026-05-27; Recorded arXiv version: v1; Recorded version updated: 2026-05-27T14:29:41Z.",
    "title": "GUI-CIDER: Mid-training GUI Agents via Causal Internalization and Density-aware Exemplar Reselection",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2605.29534v1",
    "archive": "arXiv",
    "archive_location": "2605.29534",
    "author": [
      {
        "literal": "Yuxiang Chai"
      },
      {
        "literal": "Han Xiao"
      },
      {
        "literal": "Xinyu Fu"
      },
      {
        "literal": "Jinpeng Chen"
      },
      {
        "literal": "Rui Liu"
      },
      {
        "literal": "Hongsheng Li"
      }
    ],
    "citation-key": "arxiv_2605_29534",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "461790f351a2816004027cc58d07f79ee7c670ed5064c59b1651a4efb84ff2bf",
        "submitted_at": "2026-05-28T07:49:09Z",
        "updated_at": "2026-05-28T07:49:09Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2605.29534v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2605.29534",
      "catalog_id": "arxiv:2605.29534",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-05-28",
      "identifiers": {
        "arxiv": "2605.29534"
      },
      "record_updated_at": "2026-05-28"
    },
    "genre": "preprint",
    "id": "arxiv:2605.29534",
    "issued": {
      "date-parts": [
        [
          2026,
          5,
          28
        ]
      ]
    },
    "note": "First published: 2026-05-28; Recorded arXiv version: v1; Recorded version updated: 2026-05-28T07:49:09Z.",
    "title": "UI-KOBE: Knowledge-Oriented Behavior Exploration for Lightweight Graph-Guided GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2606.01533v1",
    "archive": "arXiv",
    "archive_location": "2606.01533",
    "author": [
      {
        "literal": "Jing Yu Koh"
      },
      {
        "literal": "Ruslan Salakhutdinov"
      },
      {
        "literal": "Daniel Fried"
      }
    ],
    "citation-key": "arxiv_2606_01533",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "acc6e8af676b7de5393469d379746e674ccd662f570399622d0d5b1b0e882939",
        "submitted_at": "2026-06-01T01:29:36Z",
        "updated_at": "2026-06-01T01:29:36Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2606.01533v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2606.01533",
      "catalog_id": "arxiv:2606.01533",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-06-01",
      "identifiers": {
        "arxiv": "2606.01533"
      },
      "record_updated_at": "2026-06-01"
    },
    "genre": "preprint",
    "id": "arxiv:2606.01533",
    "issued": {
      "date-parts": [
        [
          2026,
          6,
          1
        ]
      ]
    },
    "note": "First published: 2026-06-01; Recorded arXiv version: v1; Recorded version updated: 2026-06-01T01:29:36Z.",
    "title": "Multi-Agent Computer Use",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2606.04391v1",
    "archive": "arXiv",
    "archive_location": "2606.04391",
    "author": [
      {
        "literal": "Jiaxi Li"
      },
      {
        "literal": "Ke Deng"
      },
      {
        "literal": "Yun Wang"
      },
      {
        "literal": "Jingyuan Huang"
      },
      {
        "literal": "Yucheng Shi"
      },
      {
        "literal": "Qiaoyu Tan"
      },
      {
        "literal": "Jin Lu"
      },
      {
        "literal": "Ninghao Liu"
      }
    ],
    "citation-key": "arxiv_2606_04391",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "6d891956bde9b8672859fcfcd2d54dfd3e4720964c25d77cc28bc1200ef3a56c",
        "submitted_at": "2026-06-03T03:11:50Z",
        "updated_at": "2026-06-03T03:11:50Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2606.04391v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2606.04391",
      "catalog_id": "arxiv:2606.04391",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-06-03",
      "identifiers": {
        "arxiv": "2606.04391"
      },
      "record_updated_at": "2026-06-03"
    },
    "genre": "preprint",
    "id": "arxiv:2606.04391",
    "issued": {
      "date-parts": [
        [
          2026,
          6,
          3
        ]
      ]
    },
    "note": "First published: 2026-06-03; Recorded arXiv version: v1; Recorded version updated: 2026-06-03T03:11:50Z.",
    "title": "Online Skill Learning for Web Agents via State-Grounded Dynamic Retrieval",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2606.09399v1",
    "archive": "arXiv",
    "archive_location": "2606.09399",
    "author": [
      {
        "literal": "Radeen Mostafa"
      },
      {
        "literal": "Sawradip Saha"
      }
    ],
    "citation-key": "arxiv_2606_09399",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d29dec745923e652654521f65a32f919a99d4577992006efc1a5e469bf775d81",
        "submitted_at": "2026-06-08T12:18:22Z",
        "updated_at": "2026-06-08T12:18:22Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2606.09399v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2606.09399",
      "catalog_id": "arxiv:2606.09399",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-06-08",
      "identifiers": {
        "arxiv": "2606.09399"
      },
      "record_updated_at": "2026-06-08"
    },
    "genre": "preprint",
    "id": "arxiv:2606.09399",
    "issued": {
      "date-parts": [
        [
          2026,
          6,
          8
        ]
      ]
    },
    "note": "First published: 2026-06-08; Recorded arXiv version: v1; Recorded version updated: 2026-06-08T12:18:22Z.",
    "title": "RunAgent SuperBrowser: A Theory of Autonomous Web Navigation Grounded in Human Browsing Behaviour",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2606.18890v1",
    "archive": "arXiv",
    "archive_location": "2606.18890",
    "author": [
      {
        "literal": "Zhimin Fan"
      },
      {
        "literal": "Hongwei Yu"
      },
      {
        "literal": "Yeqing Shen"
      },
      {
        "literal": "Haolong Yan"
      },
      {
        "literal": "Guozhen Peng"
      },
      {
        "literal": "Tianhao Peng"
      },
      {
        "literal": "Yudong Zhang"
      },
      {
        "literal": "Xiaowen Zhang"
      },
      {
        "literal": "Kaijun Tan"
      },
      {
        "literal": "Zheng Ge"
      },
      {
        "literal": "Xiangyu Zhang"
      },
      {
        "literal": "Daxin Jiang"
      }
    ],
    "citation-key": "arxiv_2606_18890",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "1929dfb5074e159d83006c7bbb470237a0c557e0723ac631116aa5e0369f9247",
        "submitted_at": "2026-06-17T10:07:51Z",
        "updated_at": "2026-06-17T10:07:51Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2606.18890v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2606.18890",
      "catalog_id": "arxiv:2606.18890",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-06-17",
      "identifiers": {
        "arxiv": "2606.18890"
      },
      "record_updated_at": "2026-06-17"
    },
    "genre": "preprint",
    "id": "arxiv:2606.18890",
    "issued": {
      "date-parts": [
        [
          2026,
          6,
          17
        ]
      ]
    },
    "note": "First published: 2026-06-17; Recorded arXiv version: v1; Recorded version updated: 2026-06-17T10:07:51Z.",
    "title": "Skill-Guided Continuation Distillation for GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2606.27330v1",
    "archive": "arXiv",
    "archive_location": "2606.27330",
    "author": [
      {
        "literal": "Tianyi Men"
      },
      {
        "literal": "Zhuoran Jin"
      },
      {
        "literal": "Pengfei Cao"
      },
      {
        "literal": "Yubo Chen"
      },
      {
        "literal": "Kang Liu"
      },
      {
        "literal": "Jun Zhao"
      }
    ],
    "citation-key": "arxiv_2606_27330",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "42166ac3980296a53b56bc4a076634beec36ba885b3260b5d36115b5a5261cd0",
        "submitted_at": "2026-06-25T17:44:48Z",
        "updated_at": "2026-06-25T17:44:48Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2606.27330v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2606.27330",
      "catalog_id": "arxiv:2606.27330",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-06-25",
      "identifiers": {
        "arxiv": "2606.27330"
      },
      "record_updated_at": "2026-06-25"
    },
    "genre": "preprint",
    "id": "arxiv:2606.27330",
    "issued": {
      "date-parts": [
        [
          2026,
          6,
          25
        ]
      ]
    },
    "note": "First published: 2026-06-25; Recorded arXiv version: v1; Recorded version updated: 2026-06-25T17:44:48Z.",
    "title": "Empowering GUI Agents via Autonomous Experience Exploration and Hindsight Experience Utilization for Task Planning",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2607.11388v1",
    "archive": "arXiv",
    "archive_location": "2607.11388",
    "author": [
      {
        "literal": "Wenyi Wu"
      },
      {
        "literal": "Sibo Zhu"
      },
      {
        "literal": "Kun Zhou"
      },
      {
        "literal": "Aayush Salvi"
      },
      {
        "literal": "Zixuan Song"
      },
      {
        "literal": "Biwei Huang"
      }
    ],
    "citation-key": "arxiv_2607_11388",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "40daa1b180a0c2eaba7b907a0f61b8dd8819cb3033b9da118162330d97cfecf7",
        "submitted_at": "2026-07-13T10:48:25Z",
        "updated_at": "2026-07-13T10:48:25Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2607.11388v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2607.11388",
      "catalog_id": "arxiv:2607.11388",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-07-13",
      "identifiers": {
        "arxiv": "2607.11388"
      },
      "record_updated_at": "2026-07-13"
    },
    "genre": "preprint",
    "id": "arxiv:2607.11388",
    "issued": {
      "date-parts": [
        [
          2026,
          7,
          13
        ]
      ]
    },
    "note": "First published: 2026-07-13; Recorded arXiv version: v1; Recorded version updated: 2026-07-13T10:48:25Z.",
    "title": "StructAgent: Harness Long-horizon Digital Agents with Unified Causal Structure",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2607.12625v2",
    "archive": "arXiv",
    "archive_location": "2607.12625",
    "author": [
      {
        "literal": "Yunxin Li"
      },
      {
        "literal": "Jinchao Li"
      },
      {
        "literal": "Shibo Su"
      },
      {
        "literal": "Zhenran Xu"
      },
      {
        "literal": "Chenrui Zhao"
      },
      {
        "literal": "Tongshu Bian"
      },
      {
        "literal": "Xiaoman Liang"
      },
      {
        "literal": "Meishan Zhang"
      },
      {
        "literal": "Baotian Hu"
      },
      {
        "literal": "Min Zhang"
      }
    ],
    "citation-key": "arxiv_2607_12625",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "281de2bb6e45a58efd047735accc2049deab0f8815931c86abbd38b86b6d23bf",
        "submitted_at": "2026-07-14T11:04:01Z",
        "updated_at": "2026-07-15T04:41:47Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2607.12625v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2607.12625",
      "catalog_id": "arxiv:2607.12625",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-07-14",
      "identifiers": {
        "arxiv": "2607.12625"
      },
      "record_updated_at": "2026-07-15"
    },
    "genre": "preprint",
    "id": "arxiv:2607.12625",
    "issued": {
      "date-parts": [
        [
          2026,
          7,
          14
        ]
      ]
    },
    "note": "First published: 2026-07-14; Recorded arXiv version: v2; Recorded version updated: 2026-07-15T04:41:47Z.",
    "title": "KnowAct-GUIClaw: Know Deeply, Act Perfectly, Personal GUI Assistant with Self-Evolving Memory and Skill",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2607.14548v1",
    "archive": "arXiv",
    "archive_location": "2607.14548",
    "author": [
      {
        "literal": "Hy Vision Team"
      },
      {
        "literal": "Huawen Shen"
      },
      {
        "literal": "Zhengyang Tang"
      },
      {
        "literal": "Shangpin Peng"
      },
      {
        "literal": "Liang Wu"
      },
      {
        "literal": "Anran Zhang"
      },
      {
        "literal": "Weinong Wang"
      },
      {
        "literal": "Yiduo Guo"
      },
      {
        "literal": "Chenxin Li"
      },
      {
        "literal": "Zhengyao Fang"
      },
      {
        "literal": "Yang Ding"
      },
      {
        "literal": "Junyi Li"
      },
      {
        "literal": "Fei Tang"
      },
      {
        "literal": "Zheng Ruan"
      },
      {
        "literal": "Yi Zhang"
      },
      {
        "literal": "Xingran Zhou"
      },
      {
        "literal": "Dingchen Yang"
      },
      {
        "literal": "Sunqi Fan"
      },
      {
        "literal": "Zhiyi Wan"
      },
      {
        "literal": "Han Hu"
      },
      {
        "literal": "Xin Lai"
      },
      {
        "literal": "Pengyuan Lyu"
      },
      {
        "literal": "Chengquan Zhang"
      }
    ],
    "citation-key": "arxiv_2607_14548",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "151d4882fecc835dfffb9651bd860ebe8c85919a63ff5730e191f9146fbba272",
        "submitted_at": "2026-07-16T04:12:42Z",
        "updated_at": "2026-07-16T04:12:42Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2607.14548v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2607.14548",
      "catalog_id": "arxiv:2607.14548",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-07-16",
      "identifiers": {
        "arxiv": "2607.14548"
      },
      "record_updated_at": "2026-07-16"
    },
    "genre": "preprint",
    "id": "arxiv:2607.14548",
    "issued": {
      "date-parts": [
        [
          2026,
          7,
          16
        ]
      ]
    },
    "note": "First published: 2026-07-16; Recorded arXiv version: v1; Recorded version updated: 2026-07-16T04:12:42Z.",
    "title": "HyMobileAgent: Data-Environment Co-Scaling for Efficient GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2607.15193v1",
    "archive": "arXiv",
    "archive_location": "2607.15193",
    "author": [
      {
        "literal": "Madhumitha Venkatesan"
      },
      {
        "literal": "Shicheng Wen"
      },
      {
        "literal": "Jiajing Guo"
      },
      {
        "literal": "Jorge Piazentin Ono"
      },
      {
        "literal": "Liu Ren"
      },
      {
        "literal": "Dongyu Liu"
      }
    ],
    "citation-key": "arxiv_2607_15193",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "1179c657434761187f3fa3bed99f46530a69878578b7a74c68febae20e64d6b3",
        "submitted_at": "2026-07-16T16:48:39Z",
        "updated_at": "2026-07-16T16:48:39Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2607.15193v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2607.15193",
      "catalog_id": "arxiv:2607.15193",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-07-16",
      "identifiers": {
        "arxiv": "2607.15193"
      },
      "record_updated_at": "2026-07-16"
    },
    "genre": "preprint",
    "id": "arxiv:2607.15193",
    "issued": {
      "date-parts": [
        [
          2026,
          7,
          16
        ]
      ]
    },
    "note": "First published: 2026-07-16; Recorded arXiv version: v1; Recorded version updated: 2026-07-16T16:48:39Z.",
    "title": "Plover: Steering GUI Agents through Plan-Centric Interaction",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2607.22689v1",
    "archive": "arXiv",
    "archive_location": "2607.22689",
    "author": [
      {
        "literal": "Zedong Yu"
      },
      {
        "literal": "Qianxing Li"
      },
      {
        "literal": "Zhi Gao"
      },
      {
        "literal": "Liuyu Xiang"
      },
      {
        "literal": "Chenrui Shi"
      },
      {
        "literal": "Yang Liu"
      },
      {
        "literal": "Huiming Wu"
      },
      {
        "literal": "Yujie Wei"
      },
      {
        "literal": "Yuhao Fei"
      },
      {
        "literal": "Yubo Fu"
      },
      {
        "literal": "Zhaofeng He"
      }
    ],
    "citation-key": "arxiv_2607_22689",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "078ff9a56b769923d15a98581685a3e2f47aac7d898ea731999cabaec3918fa7",
        "submitted_at": "2026-07-17T12:48:03Z",
        "updated_at": "2026-07-17T12:48:03Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2607.22689v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2607.22689",
      "catalog_id": "arxiv:2607.22689",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-07-17",
      "identifiers": {
        "arxiv": "2607.22689"
      },
      "record_updated_at": "2026-07-17"
    },
    "genre": "preprint",
    "id": "arxiv:2607.22689",
    "issued": {
      "date-parts": [
        [
          2026,
          7,
          17
        ]
      ]
    },
    "note": "First published: 2026-07-17; Recorded arXiv version: v1; Recorded version updated: 2026-07-17T12:48:03Z.",
    "title": "Beyond Sequential Interaction: Benchmarking Parallel Execution and Coordination for GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2607.24167v1",
    "archive": "arXiv",
    "archive_location": "2607.24167",
    "author": [
      {
        "literal": "Guangyi Liu"
      },
      {
        "literal": "Huan Zhao"
      },
      {
        "literal": "Quanming Yao"
      }
    ],
    "citation-key": "arxiv_2607_24167",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "57fa0a3eaad7525991656989ed3011101026362acec43895f32cf6e24be2f4ee",
        "submitted_at": "2026-07-27T08:48:01Z",
        "updated_at": "2026-07-27T08:48:01Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2607.24167v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2607.24167",
      "catalog_id": "arxiv:2607.24167",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-07-27",
      "identifiers": {
        "arxiv": "2607.24167"
      },
      "record_updated_at": "2026-07-27"
    },
    "genre": "preprint",
    "id": "arxiv:2607.24167",
    "issued": {
      "date-parts": [
        [
          2026,
          7,
          27
        ]
      ]
    },
    "note": "First published: 2026-07-27; Recorded arXiv version: v1; Recorded version updated: 2026-07-27T08:48:01Z.",
    "title": "Falsifiable Commitment Planning for Self-Correcting Web Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2607.28399v1",
    "archive": "arXiv",
    "archive_location": "2607.28399",
    "author": [
      {
        "literal": "Zihan Dong"
      },
      {
        "literal": "Rui Qian"
      },
      {
        "literal": "Qishi Zhan"
      },
      {
        "literal": "Dongshen Peng"
      },
      {
        "literal": "Kaixin Li"
      },
      {
        "literal": "Yu Li"
      }
    ],
    "citation-key": "arxiv_2607_28399",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "463cf35eb57b3d38dda6ce768e6f08d93b1c16d6abba0d790760d846cd56fa2c",
        "submitted_at": "2026-07-30T15:50:10Z",
        "updated_at": "2026-07-30T15:50:10Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2607.28399v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2607.28399",
      "catalog_id": "arxiv:2607.28399",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-07-30",
      "identifiers": {
        "arxiv": "2607.28399"
      },
      "record_updated_at": "2026-07-30"
    },
    "genre": "preprint",
    "id": "arxiv:2607.28399",
    "issued": {
      "date-parts": [
        [
          2026,
          7,
          30
        ]
      ]
    },
    "note": "First published: 2026-07-30; Recorded arXiv version: v1; Recorded version updated: 2026-07-30T15:50:10Z.",
    "title": "Why Are GUI Agents Correct but Late? Decode on the Decision-Time Critical Path, Tested with Pre-Compiled Policy Trees",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2608.00202v1",
    "archive": "arXiv",
    "archive_location": "2608.00202",
    "author": [
      {
        "literal": "Yashaswi Malla"
      },
      {
        "literal": "Sandra Siby"
      }
    ],
    "citation-key": "arxiv_2608_00202",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "60c93c90e6d4026452810a866a0c49d57afdec15ad5ced30606c925a8e83c8f7",
        "submitted_at": "2026-07-31T18:27:12Z",
        "updated_at": "2026-07-31T18:27:12Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2608.00202v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2608.00202",
      "catalog_id": "arxiv:2608.00202",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-07-31",
      "identifiers": {
        "arxiv": "2608.00202"
      },
      "record_updated_at": "2026-07-31"
    },
    "genre": "preprint",
    "id": "arxiv:2608.00202",
    "issued": {
      "date-parts": [
        [
          2026,
          7,
          31
        ]
      ]
    },
    "note": "First published: 2026-07-31; Recorded arXiv version: v1; Recorded version updated: 2026-07-31T18:27:12Z.",
    "title": "From Monoliths to Swarms: A Study of Attack Surface Evolution in the Transition to Multi-Agent Web Systems",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2608.14132v1",
    "archive": "arXiv",
    "archive_location": "2608.14132",
    "author": [
      {
        "literal": "Xiaokai Yan"
      },
      {
        "literal": "Jingtao Ding"
      },
      {
        "literal": "Yong Li"
      },
      {
        "literal": "Zhiwen Yu"
      }
    ],
    "citation-key": "arxiv_2608_14132",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "e2660e01921a78bed7280b34f35a1b587e62339c2e280d6294bbb6d6e666786e",
        "submitted_at": "2026-08-14T09:36:59Z",
        "updated_at": "2026-08-14T09:36:59Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2608.14132v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2608.14132",
      "catalog_id": "arxiv:2608.14132",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-08-14",
      "identifiers": {
        "arxiv": "2608.14132"
      },
      "record_updated_at": "2026-08-14"
    },
    "genre": "preprint",
    "id": "arxiv:2608.14132",
    "issued": {
      "date-parts": [
        [
          2026,
          8,
          14
        ]
      ]
    },
    "note": "First published: 2026-08-14; Recorded arXiv version: v1; Recorded version updated: 2026-08-14T09:36:59Z.",
    "title": "Act2Intention: A Benchmark For Developing Active Mobile Agents Through Inferring User Intention from GUI Actions",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2608.25777v2",
    "archive": "arXiv",
    "archive_location": "2608.25777",
    "author": [
      {
        "literal": "Weiming Li"
      },
      {
        "literal": "Helen Paik"
      },
      {
        "literal": "Yulei Sui"
      }
    ],
    "citation-key": "arxiv_2608_25777",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "af6b60c6a39a990e950cf4b98f905f2d3aa5ae62d4055630c25a650508ec8bf8",
        "submitted_at": "2026-08-26T13:20:27Z",
        "updated_at": "2026-09-25T23:54:04Z",
        "version": 2,
        "versioned_url": "https://arxiv.org/abs/2608.25777v2"
      },
      "canonical_source_url": "https://arxiv.org/abs/2608.25777",
      "catalog_id": "arxiv:2608.25777",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-08-26",
      "identifiers": {
        "arxiv": "2608.25777"
      },
      "record_updated_at": "2026-09-25"
    },
    "genre": "preprint",
    "id": "arxiv:2608.25777",
    "issued": {
      "date-parts": [
        [
          2026,
          8,
          26
        ]
      ]
    },
    "note": "First published: 2026-08-26; Recorded arXiv version: v2; Recorded version updated: 2026-09-25T23:54:04Z.",
    "title": "LocalLSTC: A Long Short-Term Control Architecture for Locally Deployed GUI Agents",
    "type": "article",
    "version": "2"
  },
  {
    "URL": "https://arxiv.org/abs/2608.29305v1",
    "archive": "arXiv",
    "archive_location": "2608.29305",
    "author": [
      {
        "literal": "Junxuan Li"
      },
      {
        "literal": "Zijun Liu"
      },
      {
        "literal": "Ziyi Huang"
      },
      {
        "literal": "Peng Li"
      },
      {
        "literal": "Yuzhou Liu"
      },
      {
        "literal": "Ming Yan"
      },
      {
        "literal": "Yang Liu"
      }
    ],
    "citation-key": "arxiv_2608_29305",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "bbf657b2228bad25e668e3c9d7ec20c0ecab403d34b7e0b5608d350c46b8e50f",
        "submitted_at": "2026-08-29T14:45:09Z",
        "updated_at": "2026-08-29T14:45:09Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2608.29305v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2608.29305",
      "catalog_id": "arxiv:2608.29305",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-08-29",
      "identifiers": {
        "arxiv": "2608.29305"
      },
      "record_updated_at": "2026-08-29"
    },
    "genre": "preprint",
    "id": "arxiv:2608.29305",
    "issued": {
      "date-parts": [
        [
          2026,
          8,
          29
        ]
      ]
    },
    "note": "First published: 2026-08-29; Recorded arXiv version: v1; Recorded version updated: 2026-08-29T14:45:09Z.",
    "title": "Learning Simple Test-Time Environments for LLM Web Agents",
    "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/2609.03438v1",
    "archive": "arXiv",
    "archive_location": "2609.03438",
    "author": [
      {
        "literal": "Zhaoyuan Huang"
      },
      {
        "literal": "Tianjie Ju"
      },
      {
        "literal": "Pengzhou Cheng"
      },
      {
        "literal": "Zheng Wu"
      },
      {
        "literal": "Yansi Li"
      },
      {
        "literal": "Chuanbiao Song"
      },
      {
        "literal": "Jun Lan"
      },
      {
        "literal": "Huijia Zhu"
      },
      {
        "literal": "Weiqiang Wang"
      },
      {
        "literal": "Zhuosheng Zhang"
      }
    ],
    "citation-key": "arxiv_2609_03438",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "a59f7d54845de6efebf6365a188e61e41fdf31e9ebe894b1d01ce7b86d6be4ec",
        "submitted_at": "2026-09-03T06:48:41Z",
        "updated_at": "2026-09-03T06:48:41Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2609.03438v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2609.03438",
      "catalog_id": "arxiv:2609.03438",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-09-03",
      "identifiers": {
        "arxiv": "2609.03438"
      },
      "record_updated_at": "2026-09-03"
    },
    "genre": "preprint",
    "id": "arxiv:2609.03438",
    "issued": {
      "date-parts": [
        [
          2026,
          9,
          3
        ]
      ]
    },
    "note": "First published: 2026-09-03; Recorded arXiv version: v1; Recorded version updated: 2026-09-03T06:48:41Z.",
    "title": "Do GUI Agents Know When Not to Act? Enabling Conflict-Aware Termination for Multimodal GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2609.10451v1",
    "archive": "arXiv",
    "archive_location": "2609.10451",
    "author": [
      {
        "literal": "Zixiang Chen"
      },
      {
        "literal": "Yuheng Lu"
      },
      {
        "literal": "Zihao Cheng"
      },
      {
        "literal": "Zeming Liu"
      },
      {
        "literal": "Jizeng Bai"
      },
      {
        "literal": "Ziye Huang"
      },
      {
        "literal": "Zhiyin Lin"
      },
      {
        "literal": "Zihan Li"
      },
      {
        "literal": "Yuhang Guo"
      },
      {
        "literal": "Yunhong Wang"
      },
      {
        "literal": "Haifeng Wang"
      }
    ],
    "citation-key": "arxiv_2609_10451",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "5ec17516bce87a8e349983acac14415a6ceeb12c167fce57aa37d6157cd3cd83",
        "submitted_at": "2026-09-09T16:59:59Z",
        "updated_at": "2026-09-09T16:59:59Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2609.10451v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2609.10451",
      "catalog_id": "arxiv:2609.10451",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-09-09",
      "identifiers": {
        "arxiv": "2609.10451"
      },
      "record_updated_at": "2026-09-09"
    },
    "genre": "preprint",
    "id": "arxiv:2609.10451",
    "issued": {
      "date-parts": [
        [
          2026,
          9,
          9
        ]
      ]
    },
    "note": "First published: 2026-09-09; Recorded arXiv version: v1; Recorded version updated: 2026-09-09T16:59:59Z.",
    "title": "JarvisGUI: Towards Cross-Device GUI Agents with Dynamic Task Composition",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2609.25769v1",
    "archive": "arXiv",
    "archive_location": "2609.25769",
    "author": [
      {
        "literal": "Yan Zhang"
      },
      {
        "literal": "Pei Fu"
      },
      {
        "literal": "Daiqing Wu"
      },
      {
        "literal": "Huawen Shen"
      },
      {
        "literal": "Ruoceng Zhang"
      },
      {
        "literal": "Shaojie Zhang"
      },
      {
        "literal": "Jiahui Yang"
      },
      {
        "literal": "Yu Zhou"
      },
      {
        "literal": "Can Ma"
      },
      {
        "literal": "Zhenbo Luo"
      },
      {
        "literal": "Jian Luan"
      }
    ],
    "citation-key": "arxiv_2609_25769",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "d184b111aa2a9ad65dd57391b72bcef369c2bcd01ceb8e4f00f7a2320db856d8",
        "submitted_at": "2026-09-22T07:02:35Z",
        "updated_at": "2026-09-22T07:02:35Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2609.25769v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2609.25769",
      "catalog_id": "arxiv:2609.25769",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-09-22",
      "identifiers": {
        "arxiv": "2609.25769"
      },
      "record_updated_at": "2026-09-22"
    },
    "genre": "preprint",
    "id": "arxiv:2609.25769",
    "issued": {
      "date-parts": [
        [
          2026,
          9,
          22
        ]
      ]
    },
    "note": "First published: 2026-09-22; Recorded arXiv version: v1; Recorded version updated: 2026-09-22T07:02:35Z.",
    "title": "Towards Omni-dimensional GUI Agent Navigation with Masked Trajectory Prediction",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2609.29892v1",
    "archive": "arXiv",
    "archive_location": "2609.29892",
    "author": [
      {
        "literal": "Tingyu Qu"
      },
      {
        "literal": "Weigao Sun"
      },
      {
        "literal": "Yuecheng Liu"
      },
      {
        "literal": "Yucheng Zhao"
      },
      {
        "literal": "Yi Zhu"
      },
      {
        "literal": "Yifeng Ding"
      },
      {
        "literal": "Qiyi Wang"
      },
      {
        "literal": "Sihan Cao"
      },
      {
        "literal": "Pengkun Jiao"
      },
      {
        "literal": "Hanlei Xie"
      },
      {
        "literal": "Xiongwei Wu"
      },
      {
        "literal": "Qichao Wang"
      },
      {
        "literal": "Haodong Zhang"
      },
      {
        "literal": "Jiajun Liu"
      },
      {
        "literal": "Yuhao Wang"
      },
      {
        "literal": "Yuqing Xie"
      },
      {
        "literal": "Junpeng Zhao"
      },
      {
        "literal": "Long Chen"
      },
      {
        "literal": "Ming Ma"
      },
      {
        "literal": "Sihan Yang"
      },
      {
        "literal": "Ziwang Zhao"
      },
      {
        "literal": "Yanhao Jia"
      },
      {
        "literal": "Liangquan Gong"
      },
      {
        "literal": "Feida Zhu"
      },
      {
        "literal": "Yiran Zhong"
      },
      {
        "literal": "Steven Hoi"
      }
    ],
    "citation-key": "arxiv_2609_29892",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "572ba5397fc1e4bb138febc3e7004e6e4f244e0ba8b628fb91debef4a854561f",
        "submitted_at": "2026-09-24T14:38:08Z",
        "updated_at": "2026-09-24T14:38:08Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2609.29892v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2609.29892",
      "catalog_id": "arxiv:2609.29892",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-09-24",
      "identifiers": {
        "arxiv": "2609.29892"
      },
      "record_updated_at": "2026-09-24"
    },
    "genre": "preprint",
    "id": "arxiv:2609.29892",
    "issued": {
      "date-parts": [
        [
          2026,
          9,
          24
        ]
      ]
    },
    "note": "First published: 2026-09-24; Recorded arXiv version: v1; Recorded version updated: 2026-09-24T14:38:08Z.",
    "title": "Qwen-Planner-Agent: A Closed-Loop AI-for-AI Framework for Real-World Mobile Planner Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2609.30186v1",
    "archive": "arXiv",
    "archive_location": "2609.30186",
    "author": [
      {
        "literal": "Linghua Zhang"
      }
    ],
    "citation-key": "arxiv_2609_30186",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "b387c1cfe77f6e06477415944b3923189a1f0ca30be0680f01815cb65b0067bc",
        "submitted_at": "2026-09-24T17:30:32Z",
        "updated_at": "2026-09-24T17:30:32Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2609.30186v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2609.30186",
      "catalog_id": "arxiv:2609.30186",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-09-24",
      "identifiers": {
        "arxiv": "2609.30186"
      },
      "record_updated_at": "2026-09-24"
    },
    "genre": "preprint",
    "id": "arxiv:2609.30186",
    "issued": {
      "date-parts": [
        [
          2026,
          9,
          24
        ]
      ]
    },
    "note": "First published: 2026-09-24; Recorded arXiv version: v1; Recorded version updated: 2026-09-24T17:30:32Z.",
    "title": "Jev-Mobile: Jev as an Executor for Mobile GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2609.32385v1",
    "archive": "arXiv",
    "archive_location": "2609.32385",
    "author": [
      {
        "literal": "Chuhan Zhang"
      },
      {
        "literal": "Qi Xie"
      },
      {
        "literal": "Ziyue Wang"
      },
      {
        "literal": "Jianing Yin"
      },
      {
        "literal": "Yunfan Zhou"
      },
      {
        "literal": "Dazhen Deng"
      },
      {
        "literal": "Yingcai Wu"
      }
    ],
    "citation-key": "arxiv_2609_32385",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "f091cab38c932e08d21afa936dfe241c93c0b2139aba1d1ce7a6d1513ad5da1a",
        "submitted_at": "2026-09-26T08:58:50Z",
        "updated_at": "2026-09-26T08:58:50Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2609.32385v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2609.32385",
      "catalog_id": "arxiv:2609.32385",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-09-26",
      "identifiers": {
        "arxiv": "2609.32385"
      },
      "record_updated_at": "2026-09-26"
    },
    "genre": "preprint",
    "id": "arxiv:2609.32385",
    "issued": {
      "date-parts": [
        [
          2026,
          9,
          26
        ]
      ]
    },
    "note": "First published: 2026-09-26; Recorded arXiv version: v1; Recorded version updated: 2026-09-26T08:58:50Z.",
    "title": "DashAct: A Progressive Diagnostic Benchmark for GUI Agents in Interactive Dashboard Analysis",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2609.33304v1",
    "archive": "arXiv",
    "archive_location": "2609.33304",
    "author": [
      {
        "literal": "Fuyao Zhang"
      },
      {
        "literal": "Xuan Wang"
      },
      {
        "literal": "Zherui Li"
      },
      {
        "literal": "Jiaming Zhang"
      },
      {
        "literal": "Longtao Huang"
      },
      {
        "literal": "Wei Yang Bryan Lim"
      }
    ],
    "citation-key": "arxiv_2609_33304",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "563661718e6e64d62f0601bbc5c214828219b9495a729ae685ba1ae9599332f3",
        "submitted_at": "2026-09-27T07:15:57Z",
        "updated_at": "2026-09-27T07:15:57Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2609.33304v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2609.33304",
      "catalog_id": "arxiv:2609.33304",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-09-27",
      "identifiers": {
        "arxiv": "2609.33304"
      },
      "record_updated_at": "2026-09-27"
    },
    "genre": "preprint",
    "id": "arxiv:2609.33304",
    "issued": {
      "date-parts": [
        [
          2026,
          9,
          27
        ]
      ]
    },
    "note": "First published: 2026-09-27; Recorded arXiv version: v1; Recorded version updated: 2026-09-27T07:15:57Z.",
    "title": "Relevance Does Not Imply Applicability: Experience Activation for Personal GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2609.34769v1",
    "archive": "arXiv",
    "archive_location": "2609.34769",
    "author": [
      {
        "literal": "Bingo Zhang"
      },
      {
        "literal": "Haochuan Lu"
      },
      {
        "literal": "Zongjie Li"
      },
      {
        "literal": "Genjian Li"
      },
      {
        "literal": "Ari Yu Zhang"
      },
      {
        "literal": "Chaozheng Wang"
      }
    ],
    "citation-key": "arxiv_2609_34769",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "f81fd07262dccf6783729e4336fbd77835f2fcdea030eef8dc2272e7ef82b8fc",
        "submitted_at": "2026-09-28T09:49:25Z",
        "updated_at": "2026-09-28T09:49:25Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2609.34769v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2609.34769",
      "catalog_id": "arxiv:2609.34769",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2026-09-28",
      "identifiers": {
        "arxiv": "2609.34769"
      },
      "record_updated_at": "2026-09-28"
    },
    "genre": "preprint",
    "id": "arxiv:2609.34769",
    "issued": {
      "date-parts": [
        [
          2026,
          9,
          28
        ]
      ]
    },
    "note": "First published: 2026-09-28; Recorded arXiv version: v1; Recorded version updated: 2026-09-28T09:49:25Z.",
    "title": "LongPuzzleBench: Evaluating GUI Agents on Long-Horizon Visual Puzzles",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2609.34974v1",
    "archive": "arXiv",
    "archive_location": "2609.34974",
    "author": [
      {
        "literal": "Ruozhao Yang"
      },
      {
        "literal": "Mingfei Cheng"
      },
      {
        "literal": "Xiaofei Xie"
      }
    ],
    "citation-key": "arxiv_2609_34974",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "a9bc3b57957e6c4f1af399379c4275366563b94a8b8fe01ec83e54715cfc8d11",
        "submitted_at": "2026-09-28T11:53:24Z",
        "updated_at": "2026-09-28T11:53:24Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2609.34974v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2609.34974",
      "catalog_id": "arxiv:2609.34974",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-09-28",
      "identifiers": {
        "arxiv": "2609.34974"
      },
      "record_updated_at": "2026-09-28"
    },
    "genre": "preprint",
    "id": "arxiv:2609.34974",
    "issued": {
      "date-parts": [
        [
          2026,
          9,
          28
        ]
      ]
    },
    "note": "First published: 2026-09-28; Recorded arXiv version: v1; Recorded version updated: 2026-09-28T11:53:24Z.",
    "title": "Before Acting, Change the State: Prospective State Intervention for Web Agents under Deceptive Interfaces",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2610.02542v1",
    "archive": "arXiv",
    "archive_location": "2610.02542",
    "author": [
      {
        "literal": "Dhananjay Ashok"
      },
      {
        "literal": "Shantanu Agarwal"
      },
      {
        "literal": "Vivek Datla"
      },
      {
        "literal": "Jonathan May"
      },
      {
        "literal": "Alfy Samuel"
      }
    ],
    "citation-key": "arxiv_2610_02542",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "35ded32e7bac9a875a0fa41efad43fed5fb0aab54fade5cfc2a0628b1e666587",
        "submitted_at": "2026-10-01T22:30:08Z",
        "updated_at": "2026-10-01T22:30:08Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2610.02542v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2610.02542",
      "catalog_id": "arxiv:2610.02542",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "adjacent",
      "first_published_at": "2026-10-01",
      "identifiers": {
        "arxiv": "2610.02542"
      },
      "record_updated_at": "2026-10-01"
    },
    "genre": "preprint",
    "id": "arxiv:2610.02542",
    "issued": {
      "date-parts": [
        [
          2026,
          10,
          1
        ]
      ]
    },
    "note": "First published: 2026-10-01; Recorded arXiv version: v1; Recorded version updated: 2026-10-01T22:30:08Z.",
    "title": "How To Train Your World Model: Fine-tuning vs RAG for LM-based World Modeling",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2610.03089v1",
    "archive": "arXiv",
    "archive_location": "2610.03089",
    "author": [
      {
        "literal": "Giulio Zingrillo"
      },
      {
        "literal": "Hanna Foerster"
      },
      {
        "literal": "Ilia Shumailov"
      },
      {
        "literal": "Yiren Zhao"
      },
      {
        "literal": "Robert Mullins"
      }
    ],
    "citation-key": "arxiv_2610_03089",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "7469484bb1186dd633b26e41dddc890008e893c303f93d2352f93c9d47fd78c2",
        "submitted_at": "2026-10-02T10:08:39Z",
        "updated_at": "2026-10-02T10:08:39Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2610.03089v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2610.03089",
      "catalog_id": "arxiv:2610.03089",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-10-02",
      "identifiers": {
        "arxiv": "2610.03089"
      },
      "record_updated_at": "2026-10-02"
    },
    "genre": "preprint",
    "id": "arxiv:2610.03089",
    "issued": {
      "date-parts": [
        [
          2026,
          10,
          2
        ]
      ]
    },
    "note": "First published: 2026-10-02; Recorded arXiv version: v1; Recorded version updated: 2026-10-02T10:08:39Z.",
    "title": "Securing Computer-Use Agents Against Branch Steering Attacks",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2610.04800v1",
    "archive": "arXiv",
    "archive_location": "2610.04800",
    "author": [
      {
        "literal": "Ziyang Long"
      },
      {
        "literal": "Xinqi Li"
      },
      {
        "literal": "Lujing Xing"
      },
      {
        "literal": "Hsin-Jung Yang"
      }
    ],
    "citation-key": "arxiv_2610_04800",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "baacd8b4fad7b146bb9152a54d60a617fdbed5d0142c0c0768dacea2e7f18e60",
        "submitted_at": "2026-10-03T22:56:09Z",
        "updated_at": "2026-10-03T22:56:09Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2610.04800v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2610.04800",
      "catalog_id": "arxiv:2610.04800",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-10-03",
      "identifiers": {
        "arxiv": "2610.04800"
      },
      "record_updated_at": "2026-10-03"
    },
    "genre": "preprint",
    "id": "arxiv:2610.04800",
    "issued": {
      "date-parts": [
        [
          2026,
          10,
          3
        ]
      ]
    },
    "note": "First published: 2026-10-03; Recorded arXiv version: v1; Recorded version updated: 2026-10-03T22:56:09Z.",
    "title": "MedImageOSWorld: Benchmarking GUI Agents for Medical Image Consoles",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2610.11707v1",
    "archive": "arXiv",
    "archive_location": "2610.11707",
    "author": [
      {
        "literal": "Yudong Bai"
      },
      {
        "literal": "Yihong Chen"
      },
      {
        "literal": "Quanming Yao"
      },
      {
        "literal": "Yaqing Wang"
      }
    ],
    "citation-key": "arxiv_2610_11707",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "2627960a5698d4c2f245e91ca5033be8c09e8a1ddea02490fcc1ed9f3d42217d",
        "submitted_at": "2026-10-08T11:12:57Z",
        "updated_at": "2026-10-08T11:12:57Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2610.11707v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2610.11707",
      "catalog_id": "arxiv:2610.11707",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-10-08",
      "identifiers": {
        "arxiv": "2610.11707"
      },
      "record_updated_at": "2026-10-08"
    },
    "genre": "preprint",
    "id": "arxiv:2610.11707",
    "issued": {
      "date-parts": [
        [
          2026,
          10,
          8
        ]
      ]
    },
    "note": "First published: 2026-10-08; Recorded arXiv version: v1; Recorded version updated: 2026-10-08T11:12:57Z.",
    "title": "DeltaReplay: Task-Relative Memory Reuse for Mobile GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2610.14228v1",
    "archive": "arXiv",
    "archive_location": "2610.14228",
    "author": [
      {
        "literal": "Seungjun Yu"
      },
      {
        "literal": "Hojun Choi"
      },
      {
        "literal": "Seojeong Park"
      },
      {
        "literal": "Jaeyo Shin"
      },
      {
        "literal": "Hyunjung Shim"
      }
    ],
    "citation-key": "arxiv_2610_14228",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "191e88c19124ddaddc2fed2f0e3f8e197a4c0c7f8adbc3997c056d3333754042",
        "submitted_at": "2026-10-09T18:45:11Z",
        "updated_at": "2026-10-09T18:45:11Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2610.14228v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2610.14228",
      "catalog_id": "arxiv:2610.14228",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-10-09",
      "identifiers": {
        "arxiv": "2610.14228"
      },
      "record_updated_at": "2026-10-09"
    },
    "genre": "preprint",
    "id": "arxiv:2610.14228",
    "issued": {
      "date-parts": [
        [
          2026,
          10,
          9
        ]
      ]
    },
    "note": "First published: 2026-10-09; Recorded arXiv version: v1; Recorded version updated: 2026-10-09T18:45:11Z.",
    "title": "Latent World Action Model for GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2610.16319v1",
    "archive": "arXiv",
    "archive_location": "2610.16319",
    "author": [
      {
        "literal": "Zijian Cai"
      },
      {
        "literal": "Haowei Liu"
      },
      {
        "literal": "Kailai Chen"
      },
      {
        "literal": "Lin Peng"
      },
      {
        "literal": "Cong Wan"
      },
      {
        "literal": "Ziqi Zhang"
      },
      {
        "literal": "Chunfeng Yuan"
      },
      {
        "literal": "Bing Li"
      },
      {
        "literal": "Weiming Hu"
      },
      {
        "literal": "Haiyang Xu"
      }
    ],
    "citation-key": "arxiv_2610_16319",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "4596cc60436c4fa6c2aa0951b4699b804edb73aa1b1a46900b1d14f901d20f3b",
        "submitted_at": "2026-10-12T07:42:11Z",
        "updated_at": "2026-10-12T07:42:11Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2610.16319v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2610.16319",
      "catalog_id": "arxiv:2610.16319",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-10-12",
      "identifiers": {
        "arxiv": "2610.16319"
      },
      "record_updated_at": "2026-10-12"
    },
    "genre": "preprint",
    "id": "arxiv:2610.16319",
    "issued": {
      "date-parts": [
        [
          2026,
          10,
          12
        ]
      ]
    },
    "note": "First published: 2026-10-12; Recorded arXiv version: v1; Recorded version updated: 2026-10-12T07:42:11Z.",
    "title": "HyWe: Unifying Search and Operation for Real-World Hybrid Web Tasks",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2610.17710v1",
    "archive": "arXiv",
    "archive_location": "2610.17710",
    "author": [
      {
        "literal": "Zirui Shang"
      },
      {
        "literal": "Yang Liu"
      },
      {
        "literal": "Xin Shu"
      },
      {
        "literal": "Zhi Gao"
      },
      {
        "literal": "Xinxiao Wu"
      },
      {
        "literal": "Lifeng Fan"
      }
    ],
    "citation-key": "arxiv_2610_17710",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "46ea5a1a350ccfcc12d22ab3783e05e3c580719962831561b01aeb3fcc6f6c28",
        "submitted_at": "2026-10-13T04:37:18Z",
        "updated_at": "2026-10-13T04:37:18Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2610.17710v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2610.17710",
      "catalog_id": "arxiv:2610.17710",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-10-13",
      "identifiers": {
        "arxiv": "2610.17710"
      },
      "record_updated_at": "2026-10-13"
    },
    "genre": "preprint",
    "id": "arxiv:2610.17710",
    "issued": {
      "date-parts": [
        [
          2026,
          10,
          13
        ]
      ]
    },
    "note": "First published: 2026-10-13; Recorded arXiv version: v1; Recorded version updated: 2026-10-13T04:37:18Z.",
    "title": "Adaptive-Group Process Rewards for GUI Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2610.22200v1",
    "archive": "arXiv",
    "archive_location": "2610.22200",
    "author": [
      {
        "literal": "Hanzhang Zhou"
      },
      {
        "literal": "Long Li"
      },
      {
        "literal": "Quyu Kong"
      },
      {
        "literal": "Lei Wang"
      },
      {
        "literal": "Panrong Tong"
      },
      {
        "literal": "Long Chen"
      },
      {
        "literal": "Tianyu Xia"
      },
      {
        "literal": "Gongjie Zhang"
      },
      {
        "literal": "Xu Zhang"
      },
      {
        "literal": "Chen Liu"
      },
      {
        "literal": "Jianan Zhang"
      },
      {
        "literal": "Chenglin Cai"
      },
      {
        "literal": "Zhengwen Zeng"
      },
      {
        "literal": "Steven Hoi"
      },
      {
        "literal": "Yue Wang"
      }
    ],
    "citation-key": "arxiv_2610_22200",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "beab15a2bf8a6aca19d219dbe0a047e8c6d6bd84a580df33633b5d366c175a0e",
        "submitted_at": "2026-10-15T16:51:35Z",
        "updated_at": "2026-10-15T16:51:35Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2610.22200v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2610.22200",
      "catalog_id": "arxiv:2610.22200",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-10-15",
      "identifiers": {
        "arxiv": "2610.22200"
      },
      "record_updated_at": "2026-10-15"
    },
    "genre": "preprint",
    "id": "arxiv:2610.22200",
    "issued": {
      "date-parts": [
        [
          2026,
          10,
          15
        ]
      ]
    },
    "note": "First published: 2026-10-15; Recorded arXiv version: v1; Recorded version updated: 2026-10-15T16:51:35Z.",
    "title": "OpenGCUA: Toward a Fully Open Stack for Generalist Computer Use",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2610.22280v1",
    "archive": "arXiv",
    "archive_location": "2610.22280",
    "author": [
      {
        "literal": "Qingni Wang"
      },
      {
        "literal": "Changdae Oh"
      },
      {
        "literal": "Yuzhe Yang"
      },
      {
        "literal": "Sharon Li"
      },
      {
        "literal": "Jiachen Yang"
      },
      {
        "literal": "Ang Li"
      },
      {
        "literal": "Xin Eric Wang"
      }
    ],
    "citation-key": "arxiv_2610_22280",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "6a249a56a61ebb14b456500f9b245efe7b72997af5d5f515e5cc7707fe68b018",
        "submitted_at": "2026-10-15T17:40:12Z",
        "updated_at": "2026-10-15T17:40:12Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2610.22280v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2610.22280",
      "catalog_id": "arxiv:2610.22280",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-10-15",
      "identifiers": {
        "arxiv": "2610.22280"
      },
      "record_updated_at": "2026-10-15"
    },
    "genre": "preprint",
    "id": "arxiv:2610.22280",
    "issued": {
      "date-parts": [
        [
          2026,
          10,
          15
        ]
      ]
    },
    "note": "First published: 2026-10-15; Recorded arXiv version: v1; Recorded version updated: 2026-10-15T17:40:12Z.",
    "title": "LEASH: Uncertainty-Aware Weak-to-Strong Escalation and Oversight for Computer-Use Agents",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2610.22697v1",
    "archive": "arXiv",
    "archive_location": "2610.22697",
    "author": [
      {
        "literal": "Umar Iqbal"
      }
    ],
    "citation-key": "arxiv_2610_22697",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "077dd623c980813e83197a38b6545161a3d2fc842d99fe291c53372b9770acd2",
        "submitted_at": "2026-10-15T22:02:45Z",
        "updated_at": "2026-10-15T22:02:45Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2610.22697v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2610.22697",
      "catalog_id": "arxiv:2610.22697",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-10-15",
      "identifiers": {
        "arxiv": "2610.22697"
      },
      "record_updated_at": "2026-10-15"
    },
    "genre": "preprint",
    "id": "arxiv:2610.22697",
    "issued": {
      "date-parts": [
        [
          2026,
          10,
          15
        ]
      ]
    },
    "note": "First published: 2026-10-15; Recorded arXiv version: v1; Recorded version updated: 2026-10-15T22:02:45Z.",
    "title": "What It Is, Not What It Says: Securing Agentic Browsing Interactions with Web Semantics",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://arxiv.org/abs/2610.23012v1",
    "archive": "arXiv",
    "archive_location": "2610.23012",
    "author": [
      {
        "literal": "Chulhoon Yoon"
      },
      {
        "literal": "Kyungri Park"
      },
      {
        "literal": "Jinyeong Yun"
      },
      {
        "literal": "Daeyun Jeong"
      },
      {
        "literal": "Sukmin Yun"
      },
      {
        "literal": "Woohwan Jung"
      }
    ],
    "citation-key": "arxiv_2610_23012",
    "custom": {
      "arxiv_snapshot": {
        "abstract_hash_method": "sha256-utf8-whitespace-collapsed-v1",
        "abstract_sha256": "b9585696d256bacc710dce57bfd7cda6a8037ce5725a30f959028352bc07e97c",
        "submitted_at": "2026-10-16T06:44:56Z",
        "updated_at": "2026-10-16T06:44:56Z",
        "version": 1,
        "versioned_url": "https://arxiv.org/abs/2610.23012v1"
      },
      "canonical_source_url": "https://arxiv.org/abs/2610.23012",
      "catalog_id": "arxiv:2610.23012",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "first_published_at": "2026-10-16",
      "identifiers": {
        "arxiv": "2610.23012"
      },
      "record_updated_at": "2026-10-16"
    },
    "genre": "preprint",
    "id": "arxiv:2610.23012",
    "issued": {
      "date-parts": [
        [
          2026,
          10,
          16
        ]
      ]
    },
    "note": "First published: 2026-10-16; Recorded arXiv version: v1; Recorded version updated: 2026-10-16T06:44:56Z.",
    "title": "Adaptation of Computer Use Agents to New Software with User Manuals",
    "type": "article",
    "version": "1"
  },
  {
    "URL": "https://aaai.org/Papers/AAAI/2007/AAAI07-240.pdf",
    "author": [
      {
        "literal": "James Allen"
      },
      {
        "literal": "Nathanael Chambers"
      },
      {
        "literal": "George Ferguson"
      },
      {
        "literal": "Lucian Galescu"
      },
      {
        "literal": "Hyuckchul Jung"
      },
      {
        "literal": "Mary Swift"
      },
      {
        "literal": "William Taysom"
      }
    ],
    "citation-key": "cua_4e2f12a215bd1cadd864f498",
    "container-title": "Twenty-Second AAAI National Conference on Artificial Intelligence",
    "custom": {
      "canonical_source_url": "https://aaai.org/Papers/AAAI/2007/AAAI07-240.pdf",
      "catalog_id": "url:628b5251757bb3888e89",
      "catalog_kind": "paper",
      "catalog_review_status": "curated",
      "catalog_scope": "core",
      "citation_edition": "publication",
      "first_published_at": null,
      "identifiers": {},
      "publication": {
        "authors": [
          "James Allen",
          "Nathanael Chambers",
          "George Ferguson",
          "Lucian Galescu",
          "Hyuckchul Jung",
          "Mary Swift",
          "William Taysom"
        ],
        "date_basis": "Supplied year precision: 2007. Papers section: AAAI-2007",
        "date_parts": [
          2007
        ],
        "doi": null,
        "event_dates": null,
        "evidence_urls": [
          "https://aaai.org/Papers/AAAI/2007/AAAI07-240.pdf",
          "https://www.cs.rochester.edu/research/cisd/projects/plow/"
        ],
        "identity_basis": "No same-work record located in the full active catalog. Similarity screening was not used as automatic identity approval. The raw archive and exclusions are additionally checked during normalization.",
        "issue": null,
        "notes": "The ordered names follow the original published paper, without expanding initials or replacing names from secondary indexes. DOI and volume number are not supplied in the retained primary sources. No exact publication month/day is established.",
        "pages": "1514-1519",
        "pdf_url": "https://cdn.aaai.org/AAAI/2007/AAAI07-240.pdf",
        "publisher": null,
        "series": null,
        "title": "PLOW: A Collaborative Task Learning Agent",
        "type": "conference-paper",
        "url": "https://aaai.org/Papers/AAAI/2007/AAAI07-240.pdf",
        "venue": "Twenty-Second AAAI National Conference on Artificial Intelligence",
        "volume": null
      },
      "record_updated_at": null
    },
    "id": "url:628b5251757bb3888e89",
    "issued": {
      "date-parts": [
        [
          2007
        ]
      ]
    },
    "note": "Publication date: 2007.",
    "page": "1514-1519",
    "title": "PLOW: A Collaborative Task Learning Agent",
    "type": "paper-conference"
  }
]