Paper Detail

Exclusive Unlearning

Mutsumi Sasaki, Kouta Nakayama, Yusuke Miyao, Yohei Oseki, Masaru Isonuma

arxiv Score 6.5

Published 2026-04-07 · First seen 2026-04-08

Research Track A · General AI

Abstract

When introducing Large Language Models (LLMs) into industrial applications, such as healthcare and education, the risk of generating harmful content becomes a significant challenge. While existing machine unlearning methods can erase specific harmful knowledge and expressions, diverse harmful content makes comprehensive removal difficult. In this study, instead of individually listing targets for forgetting, we propose Exclusive Unlearning (EU), which aims for broad harm removal by extensively forgetting everything except for the knowledge and expressions we wish to retain. We demonstrate that through Exclusive Unlearning, it is possible to obtain a model that ensures safety against a wide range of inputs, including jailbreaks, while maintaining the ability to respond to diverse instructions related to specific domains such as medicine and mathematics.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
soon
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@article{sasaki2026exclusive,
  title = {Exclusive Unlearning},
  author = {Mutsumi Sasaki and Kouta Nakayama and Yusuke Miyao and Yohei Oseki and Masaru Isonuma},
  year = {2026},
  abstract = {When introducing Large Language Models (LLMs) into industrial applications, such as healthcare and education, the risk of generating harmful content becomes a significant challenge. While existing machine unlearning methods can erase specific harmful knowledge and expressions, diverse harmful content makes comprehensive removal difficult. In this study, instead of individually listing targets for forgetting, we propose Exclusive Unlearning (EU), which aims for broad harm removal by extensively f},
  url = {https://arxiv.org/abs/2604.06154},
  keywords = {cs.CL},
  eprint = {2604.06154},
  archiveprefix = {arXiv},
}

Metadata

{}