README.md

---
license: mit
tags:
- protein-structure-prediction
- biomolecular-modeling
- structural-biology
- computational-biology
- alphafold
pipeline_tag: other
---

# Triangle Multiplication is All You Need for Biomolecular Structure Representations

By [Jeffrey Ouyang-Zhang](https://jozhang97.github.io/), [Pranav Murugan](https://pranavmurugan.github.io), [Daniel J. Diaz](http://danny305.github.io), [Gianluca Scarpellini](https://scarpellini.dev/), [Richard Strong Bowen](http://rsbowen.github.io), [Nate Gruver](https://ngruver.github.io), [Adam Klivans](https://www.cs.utexas.edu/users/klivans/), [Philipp Krähenbühl](http://www.philkr.net/), [Aleksandra Faust](https://www.afaust.info), and [Maruan Al-Shedivat](https://maruan.alshedivat.com)

*Genesis Research & UT Austin*

Please see our [GitHub Repository](https://github.com/genesistherapeutics/pairmixer) for instructions on how to use this model.

## Model Description

Pairmixer is an efficient alternative to AlphaFold3's Pairformer architecture for biomolecular structure prediction. By eliminating triangle attention while preserving geometric reasoning through triangle multiplication, Pairmixer achieves:

- **4× faster inference** compared to Pairformer
- **34% lower training cost**
- Competitive accuracy on standard benchmarks

## Citation

If you use Pairmixer in your research, please cite:

```bibtex
@inproceedings{pairmixer2025,
  title={Triangle Multiplication is All You Need for Biomolecular Structure Representations},
  author={Ouyang-Zhang, Jeffrey and Murugan, Pranav and Diaz, Daniel J. and Scarpellini, Gianluca and Bowen, Richard Strong and Gruver, Nate and Klivans, Adam and Kr{\"a}henb{\"u}hl, Philipp and Faust, Aleksandra and Al-Shedivat, Maruan},
  year={2025},
}
```

## Additional Resources

- [Paper](https://arxiv.org/abs/2510.18870) (ArXiv)
- [GitHub Repository](https://github.com/genesistherapeutics/pairmixer)

## License

This model is released under the MIT License.