Official implementation of the paper " Connecting Neural Models Latent Geometries with Relative Geodesic Representations." by Hanlin Yu, Berfin Inal, Georgios Arvanitidis, Soren Hauberg, Francesco Locatello, Marco Fumero. Published at The Thirty-ninth Annual Conference on Neural Information Processing Systems, NeurIPS, 2024. This repository offers a lightweight, standalone implementation of the framework introduced in the paper.
Install the dependencies via:
cd vision_foundation_models
pip install requirements.txt
In the vision_foundation_models folder you will find instructions on how to run our experiments.
You're good to go!
If you find our code or paper useful, please cite it at:
@inproceedings{
yu2025connecting,
title={Connecting Neural Models Latent Geometries with Relative Geodesic Representations},
author={Hanlin Yu and Berfin Inal and Georgios Arvanitidis and S{\o}ren Hauberg and Francesco Locatello and Marco Fumero},
booktitle={The Thirty-ninth Annual Conference on Neural Information Processing Systems},
year={2025},
url={https://openreview.net/forum?id=4JnZvkVssS}
}