# New open source automata project

**URL:** <https://discuss.ai.google.dev/t/new-open-source-automata-project/29585>\
**Category:** TensorFlow\
**Tags:** help\_request\
**Created:** [November 5, 2022, 6:42pm UTC](https://discuss.ai.google.dev/t/new-open-source-automata-project/29585 "2022-11-05T18:42:24Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Avery\_Thomas](https://avatars.discourse-cdn.com/v4/letter/a/e56c9b/32.png) [@Avery\_Thomas](https://discuss.ai.google.dev/u/Avery_Thomas)\
**Post date:** [November 5, 2022, 6:42pm UTC](https://discuss.ai.google.dev/t/new-open-source-automata-project/29585/1 "2022-11-05T18:42:24Z")

</div>

I’m looking to fork a TensorFlow project for research but need help from experts. The new open source project would involve taking this existing project and making a few changes, I just don’t have a coding background: [GitHub - distillpub/post--growing-ca: Growing Neural Cellular Automata](https://github.com/distillpub/post--growing-ca)

Background article: [Growing Neural Cellular Automata](https://distill.pub/2020/growing-ca/)

Instead of using a starting image the input data would take the first time step and use that as the input for the next time step and so on. The output image would be a graphical representation of the stacked time changes of all alive cells.

I’m hoping the result generates a 3D hypergraph that has the same qualities as the original GitHub project like self healing.

I want to see what the algorithm does when parts are corrupted or removed. My theory is the neural network will have some interesting characteristics.