What it is
ALFWorld connects text-based household environments with corresponding embodied tasks. It supports studying agents that learn task structure through interactive language-based experience.
Text interaction offers a way to study planning and task execution before transferring those skills to more complex embodied settings.
How an agent works in it
Sees
Text descriptions of the environment, or observations from the embodied setting.
Does
Navigate, inspect objects and perform household interactions.
Feedback
Environment task progress and completion feedback.
Example task (illustrative): Find an object, prepare it, and put it in the requested location.
Setup
Install the environment and download the required task assets. Choose the text or embodied backend using the upstream configuration.
Use it with BoltzLabs
BoltzLabs does not host this project. To record an agent working on one of its tasks:
- Prepare one task from the project: its starter files, instructions, dependencies and check command.
- In Evals, choose New eval, upload the starter folder or enter its repository, then set the task and check command.
- Choose BoltzLabs models or your own endpoint. Each run gets its own machine.
- When the run finishes, open it to read the agent’s trace, file changes and check result.
Projects that need a desktop or phone, self-hosted websites, a game or a simulator need that runtime set up separately; an eval machine does not include it. Machine time uses account credits; hosted model calls are billed separately.