This is the code repository of the following paper on task-incremental lifelong learning and unlearning by leveraging task-specific sparse subnetworks with knowledge transfer.
"Privacy-Aware Lifelong Learning"
Ozan Özdenizci, Elmar Rueckert, Robert Legenstein
International Conference on Learning Representations (ICLR), 2025.
https://openreview.net/forum?id=UstOpZCESc
If you use this code or models in your research and find it helpful, please cite the following paper:
@inproceedings{ozdenizci2025pall,
title={Privacy-aware lifelong learning},
author={Ozan {\"O}zdenizci and Elmar Rueckert and Robert Legenstein},
booktitle={International Conference on Learning Representations (ICLR)},
year={2025},
url={https://openreview.net/forum?id=UstOpZCESc}
}
This work was supported by the Graz Center for Machine Learning (GraML). This research was funded in whole or in part by the Austrian Science Fund (FWF) [10.55776/COE12].