VM

Selected research

All research

Paper explained

arXiv Undone CS 2026 Machine unlearning

Memory Undone: Between Knowing and Not Knowing in Data Systems

Viktoriia Makovska, George Fletcher, Julia Stoyanovich, Tetiana Zakharchenko

What disappears when a data system “forgets”—and who decides?

The question

Why this work

This position paper treats forgetting as both a technical intervention and a social practice. Removing a file, changing a trained model, and never collecting a perspective can all make information unavailable, but through different mechanisms.

The approach

A visual guide

Four terms that should not be collapsed into one

  1. Erasure

    Remove or disable access to stored artifacts: files, rows, caches.

  2. Unlearning

    Intervene on learned parameters or outputs to remove or bound data influence.

  3. Exclusion

    Choose upstream what not to collect, label, or include.

  4. Forgetting

    The broader sociotechnical process through which something becomes unavailable.

Conceptual distinctions from the paper. These categories describe different mechanisms; they are not consecutive steps or measured outcomes. Pages 1–3

What the work contributes

The paper offers a vocabulary for separating these mechanisms and argues for accountability around what is retained or excluded. Evaluating forgetting therefore involves rights, power, and representation alongside technical utility.

Pages 1–3

Original paper & resources

arXiv preprint arXiv:2602.21180 [cs.CY], submitted February 24, 2026. Presented at Undone Computer Science on 24 March 2026.

Citation
Makovska, V., Fletcher, G., Stoyanovich, J., & Zakharchenko, T. (2026). Memory Undone: Between Knowing and Not Knowing in Data Systems. arXiv preprint arXiv:2602.21180. https://arxiv.org/abs/2602.21180
BibTeX
@misc{Makovska2026MemoryUndone,
  title = {Memory Undone: Between Knowing and Not Knowing in Data Systems},
  author = {Makovska, Viktoriia and Fletcher, George and Stoyanovich, Julia and Zakharchenko, Tetiana},
  year = {2026},
  eprint = {2602.21180},
  archivePrefix = {arXiv},
  primaryClass = {cs.CY},
  url = {https://arxiv.org/abs/2602.21180}
}