Skip to content

added ICLR bibtex info #1

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
eleniTriantafillou merged 1 commit into from
Feb 4, 2026
Merged

added ICLR bibtex info #1

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
16 changes: 15 additions & 1 deletion README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@

# Redirection for Erasing Memory (REM): Towards a universal unlearning method for corrupted data
# Redirection for Erasing Memory (REM): Towards a universal unlearning method for corrupted data [ICLR 2026]
*[Stefan Schoepf](https://if-loops.github.io/) (2)°, [Michael Curtis Mozer](https://home.cs.colorado.edu/~mozer/index.php) (1), [Nicole Elyse Mitchell](https://nicolemitchell.github.io/) (3), [Alexandra Brintrup](https://www.ifm.eng.cam.ac.uk/people/ab702/) (2), [Georgios Kaissis](https://www.g-k.ai/) (1), [Peter Kairouz](https://kairouzp.github.io/) (3), [Eleni Triantafillou](https://elenitriantafillou.github.io/) (1)*
°Work done during the author’s internship at (1) Google DeepMind; (2) University of Cambridge; (3) Google Research
> Machine unlearning is studied for a multitude of tasks, but specialization of
Expand Down Expand Up @@ -39,6 +39,19 @@ If desired, plotting.py and tables.py can be called manually.
The code in this repo builds upon the repo of [Goel et al. ("Corrective Unlearning")](https://github.com/drimpossible/corrective-unlearning-bench)(GNU GENERAL PUBLIC LICENSE) with modifications only made where necessary to enable additional functionality (e.g., adding ETD training and additional methods such as REM and Gradient Ascent) or to improve user experience (e.g., Experiments_run.sh which allows quick adjustment of experiments instead of the original hardcoded experiment list).
Additional method implementations are taken from the respective original author repositories where available (e.g., [ETD](https://github.com/pratyushmaini/localizing-memorization), [Potion](https://github.com/if-loops/towards_poison_unlearning), ...)
## Citing this work

ICLR 2026: [OpenReview](https://openreview.net/forum?id=xG0mQ4Xsfm)
```
@inproceedings{
schoepf2026redirection,
title={Redirection for Erasing Memory ({REM}): Towards a universal unlearning method for corrupted data},
author={Stefan Schoepf and Michael Curtis Mozer and Nicole Elyse Mitchell and Alexandra Brintrup and Georgios Kaissis and Peter Kairouz and Eleni Triantafillou}},
booktitle={The Fourteenth International Conference on Learning Representations},
year={2026},
url={https://openreview.net/forum?id=xG0mQ4Xsfm}
}
```

Preprint: [ArXiv](https://arxiv.org/abs/2505.17730)
```
@misc{schoepf2025remunlearning,
Expand All @@ -50,6 +63,7 @@ Preprint: [ArXiv](https://arxiv.org/abs/2505.17730)
primaryClass={cs.LG},
url={https://arxiv.org/abs/2505.17730}}
```

## License and disclaimer
Copyright 2025 Google LLC
All software is distributed under the terms of the GNU GENERAL PUBLIC LICENSE,
Expand Down