Heimdall

Heimdall#


Heimdall is a Python toolkit for implementing and benchmarking tokenizers for single-cell foundation models. Please cite our preprint if you use Heimdall in your work:
@article{haber2025heimdall,
  title={HEIMDALL: A Modular Framework for Tokenization in Single-Cell Foundation Models},
  author={Haber, Ellie and Alam, Shahul and Ho, Nicholas and Liu, Renming and Trop, Evan and Liang, Shaoheng and Yang, Muyu and Krieger, Spencer and Ma, Jian},
  journal={bioRxiv},
  pages={2025--11},
  year={2025},
  publisher={Cold Spring Harbor Laboratory}
}

Check out the documentation links below for installation instructions, usage examples, and the complete API.

Installation
installation.html
Usage
usage.html