# VEM GUI-Agent-RL

Training code that learns a value environment model from offline Android interaction data and uses its frozen value estimates to guide GUI policy optimization.

- Catalog URL: https://awesomecua.github.io/resources/github-microsoft-gui-agent-rl/
- Original source: https://github.com/microsoft/GUI-Agent-RL
- Kind: framework
- Review: curated
- Scope: core
- Environments: android
- Topics: reinforcement-learning, reward-models, mobile-navigation
- Evidence checked: 2026-10-17T09:12:14Z
- ID: github:microsoft/gui-agent-rl

## Limitations

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.

- Record JSON: https://awesomecua.github.io/resources/github-microsoft-gui-agent-rl/record.json

## License declarations

- code: MIT — https://raw.githubusercontent.com/microsoft/GUI-Agent-RL/HEAD/LICENSE
- data: not recorded
- document: not recorded
- weights: not recorded

- code review: Repository LICENSE grants MIT for code; linked training imagery and pretrained model licenses remain separate.
  Sources: https://raw.githubusercontent.com/microsoft/GUI-Agent-RL/HEAD/LICENSE

## Primary evidence

- https://raw.githubusercontent.com/microsoft/GUI-Agent-RL/HEAD/README.md (retrieved 2026-10-17T09:12:14Z; SHA-256 d21a1ae4359296c0f3034ac655888dce43448d10afe0b1909d6b6520678c2526)
- https://api.github.com/repos/microsoft/GUI-Agent-RL (retrieved 2026-10-17T09:12:13Z; SHA-256 791c6e56e62ce5e803a4ba313557bac0a39dfeba27499a3865ad444a860b426a)
- https://api.github.com/repos/microsoft/GUI-Agent-RL/git/trees/main?recursive=1 (retrieved 2026-10-17T09:14:06Z; SHA-256 5b799a5bdae78ff50652c0b464b6cad12d9dda265d4752e5d5f9eab65e9d5c6f)
- https://raw.githubusercontent.com/microsoft/GUI-Agent-RL/HEAD/LICENSE (retrieved 2026-10-17T09:14:06Z; SHA-256 c2cfccb812fe482101a8f04597dfc5a9991a6b2748266c47ac91b6a5aae15383)