metadata

title: Neural Computer
emoji: 🧠
colorFrom: purple
colorTo: blue
sdk: docker
pinned: false

Neural Computer Demo

This is a demonstration of a Neural Computer system that can generate computer screen interactions in real-time. The system uses a trained diffusion model to predict what the screen should look like based on mouse movements, clicks, and keyboard inputs. For more details, see https://arxiv.org/abs/2507.08800v1.

How to Use

Wait for the model to load - This may take a minute or two on first startup
Click anywhere on the canvas to begin interacting
Move your mouse around to see the model predict screen changes
Click and drag to simulate mouse interactions
Use keyboard inputs while focused on the canvas
Use the controls to:
- Reset the simulation
- Adjust sampling steps (lower = faster, higher = better quality)
- Toggle RNN mode for even faster inference

Settings

Sampling Steps: Controls the quality vs speed tradeoff (1-50 steps)
Use RNN: Enables faster inference mode using RNN output directly
Reset: Clears the simulation and starts fresh

Technical Details

This system uses a specialized diffusion model trained on computer interaction data. The model can predict realistic screen changes based on user inputs in real-time.

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

Citation

@misc{rivard2025neuralossimulatingoperatingsystems,
      title={NeuralOS: Towards Simulating Operating Systems via Neural Generative Models}, 
      author={Luke Rivard and Sun Sun and Hongyu Guo and Wenhu Chen and Yuntian Deng},
      year={2025},
      eprint={2507.08800},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2507.08800}, 
}