Skip to content

Official code for NeurIPS 2024 paper "A Canonicalization Perspective on Invariant and Equivariant Learning".

License

Notifications You must be signed in to change notification settings

PKU-ML/canonicalization

Repository files navigation

A Canonicalization Perspective on Invariant and Equivariant Learning

Source code for the paper "A Canonicalization Perspective on Invariant and Equivariant Learning", NeurIPS 2024.

Attribution: Our code is built on top of the [SignNet repo] by Lim et al. in 2022, which in turn builds off of the setup in [LSPE repo] by Dwivedi et al. in 2021.

If you use our code, please cite

@inproceedings{canonicalization-perspective,
    title={{A Canonicalization Perspective on Invariant and Equivariant Learning}},
    author={Ma, George and Wang, Yifei and Lim, Derek and Jegelka, Stefanie and Wang, Yisen},
    booktitle={NeurIPS},
    year={2024}
}

Releases

No releases published

Packages

No packages published