2601.06301v1.pdf

{
  "dorsal/arxiv": [
    {
      "date_modified": "2026-02-17T05:53:07.963000Z",
      "hash": "e52e05f1fd7bae58c46f59848c13df16164b233d131ec38538f8e36a9e018004",
      "id": "00d2df5b-84a5-4ee9-94b8-dd0652f40043",
      "source": {
        "execution_id": "01da3ca2-617f-46a1-8086-8d81aa020bd9",
        "id": "arXiv Dataset",
        "type": "Model",
        "variant": "snapshot-2026-01-17",
        "version": "0.1.0"
      },
      "user_id": 1000002
    }
  ],
  "file/base": {
    "record": {
      "extension": ".pdf",
      "hash": "e52e05f1fd7bae58c46f59848c13df16164b233d131ec38538f8e36a9e018004",
      "media_type": "application/pdf",
      "media_type_prefix": "application",
      "name": "2601.06301v1.pdf",
      "size": 558503
    },
    "source": {
      "execution_id": "dc420da3-9fd9-4c69-ae5c-e83a1bccf662",
      "id": "file/base",
      "type": "Model",
      "version": "1.0.0"
    }
  },
  "file/pdf": {
    "private": false,
    "record": {
      "keywords": [],
      "page_count": 9,
      "producer": "pikepdf 8.15.1",
      "title": "Beyond BeautifulSoup: Benchmarking LLM-Powered Web Scraping for Everyday Users",
      "version": "1.7"
    },
    "source": {
      "execution_id": "59c018e4-c815-4138-b221-881aeea5e0e7",
      "id": "dorsal/pdf",
      "type": "Model",
      "version": "1.1.0"
    }
  }
}