wooyeolbaek/attention-map-diffusers

🚀 Cross attention map tools for huggingface/diffusers

59
/ 100
Established

This tool helps AI researchers and practitioners understand how large language models interpret text prompts when generating images. It takes a text prompt and an image generation model (like Stable Diffusion 3) as input. It then visually highlights which parts of the input text prompt influence specific regions of the generated image, outputting these as attention maps. This is useful for debugging model behavior or fine-tuning creative outputs.

397 stars. Available on PyPI.

Use this if you need to visualize the internal 'thinking process' of a text-to-image model to see how different words in your prompt contribute to specific elements in the generated image.

Not ideal if you are looking for a simple image generation tool without needing to delve into the underlying model mechanics.

AI research text-to-image generation model interpretability prompt engineering generative AI
Maintenance 10 / 25
Adoption 10 / 25
Maturity 25 / 25
Community 14 / 25

How are scores calculated?

Stars

397

Forks

28

Language

Python

License

MIT

Last pushed

Feb 02, 2026

Commits (30d)

0

Dependencies

7

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/wooyeolbaek/attention-map-diffusers"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.