Implementation of Li et al, "Measuring the intrinsic dimension of objective landscapes"
This is an attempt to reproduce the results from the above paper in pytorch. The code is adapted from the authors' implementation with certain modifications.
To cite this:
@misc{rojagtap_intrinsic_dimension_2024,
author = {Rohan Jagtap},
title = {Measuring the Intrinsic Dimension of Objective Landscapes},
year = {2024},
url = {https://github.com/rojagtap/intrinsic-dimension-of-deep-learning-tasks},
note = {GitHub repository}
}
@inproceedings{li_id_2018_ICLR
title={Measuring the Intrinsic Dimension of Objective Landscapes},
author={Li, Chunyuan and Farkhoor, Heerad and Liu, Rosanne and Yosinski, Jason},
booktitle={International Conference on Learning Representations},
year={2018}
}
or their implementation https://github.com/uber-research/intrinsic-dimension