{
  "aliases": [],
  "authors": [],
  "details": {
    "availability": "Public training and evaluation source; trained checkpoints are generated by the documented workflow.",
    "license_review": {
      "code": {
        "evidence_urls": [
          "https://raw.githubusercontent.com/microsoft/GUI-Agent-RL/HEAD/LICENSE"
        ],
        "reason": "Repository LICENSE grants MIT for code; linked training imagery and pretrained model licenses remain separate."
      }
    },
    "repository_revision": "6d2381ef549d980452a828471221a2ce8023bf41"
  },
  "environments": [
    "android"
  ],
  "evidence": [
    {
      "method": "editorial_primary_review",
      "retrieved_at": "2026-10-17T09:12:14Z",
      "role": "primary",
      "sha256": "d21a1ae4359296c0f3034ac655888dce43448d10afe0b1909d6b6520678c2526",
      "supports": [
        "name",
        "kind",
        "summary",
        "environments",
        "topics",
        "scope",
        "notes",
        "relations",
        "details"
      ],
      "url": "https://raw.githubusercontent.com/microsoft/GUI-Agent-RL/HEAD/README.md"
    },
    {
      "method": "editorial_primary_review",
      "retrieved_at": "2026-10-17T09:12:13Z",
      "role": "primary",
      "sha256": "791c6e56e62ce5e803a4ba313557bac0a39dfeba27499a3865ad444a860b426a",
      "supports": [
        "identity"
      ],
      "url": "https://api.github.com/repos/microsoft/GUI-Agent-RL"
    },
    {
      "method": "editorial_primary_review",
      "retrieved_at": "2026-10-17T09:14:06Z",
      "role": "primary",
      "sha256": "5b799a5bdae78ff50652c0b464b6cad12d9dda265d4752e5d5f9eab65e9d5c6f",
      "supports": [
        "release_status"
      ],
      "url": "https://api.github.com/repos/microsoft/GUI-Agent-RL/git/trees/main?recursive=1"
    },
    {
      "method": "editorial_primary_review",
      "retrieved_at": "2026-10-17T09:14:06Z",
      "role": "primary",
      "sha256": "c2cfccb812fe482101a8f04597dfc5a9991a6b2748266c47ac91b6a5aae15383",
      "supports": [
        "licenses.code"
      ],
      "url": "https://raw.githubusercontent.com/microsoft/GUI-Agent-RL/HEAD/LICENSE"
    }
  ],
  "id": "github:microsoft/gui-agent-rl",
  "identifiers": {},
  "kind": "framework",
  "last_verified_at": "2026-10-17T09:12:14Z",
  "licenses": {
    "code": {
      "basis": "Repository LICENSE grants MIT for code; linked training imagery and pretrained model licenses remain separate.",
      "declared": "MIT",
      "retrieved_at": "2026-10-17T09:14:06Z",
      "url": "https://raw.githubusercontent.com/microsoft/GUI-Agent-RL/HEAD/LICENSE"
    },
    "data": null,
    "document": null,
    "weights": null
  },
  "links": [
    {
      "rel": "primary",
      "url": "https://github.com/microsoft/GUI-Agent-RL"
    }
  ],
  "name": "VEM GUI-Agent-RL",
  "notes": "The public tree includes critic-training configurations, policy training and offline/online evaluation scripts. The README obtains base models and AITW imagery separately, then instructs users to train critic and policy checkpoints locally; this entry does not claim that VEM-trained weights are released. MIT applies to repository code only.",
  "published_at": null,
  "relations": [
    {
      "evidence_url": "https://raw.githubusercontent.com/microsoft/GUI-Agent-RL/HEAD/README.md",
      "target_id": "arxiv:2502.18906",
      "type": "documented_by"
    }
  ],
  "review_status": "curated",
  "scope": "core",
  "summary": "Training code that learns a value environment model from offline Android interaction data and uses its frozen value estimates to guide GUI policy optimization.",
  "topics": [
    "reinforcement-learning",
    "reward-models",
    "mobile-navigation"
  ],
  "updated_at": null,
  "url": "https://github.com/microsoft/GUI-Agent-RL"
}