Metraipolis › archive › dc_7df2ff8ff26c

Adaptive Localization of Knowledge Negation for Continual LLM Unlearning — arXiv:2611.02620

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.02620
License (as archived): CC0-1.0
Archived: 2026-11-03T11:22:49.716Z
SHA-256 of the archived text: 607f3f23cd371370f95952f3802c5a106a78cdb52e9fcb0ef95cb86ff167534e
Stable URL of this page: https://metraipolis.com/documento/dc_7df2ff8ff26c

How to cite

Adaptive Localization of Knowledge Negation for Continual LLM Unlearning — arXiv:2611.02620 — arXiv (en), https://arxiv.org/abs/2611.02620 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7df2ff8ff26c (Metraipolis archive dc_7df2ff8ff26c, sha256 607f3f23cd371370f95952f3802c5a106a78cdb52e9fcb0ef95cb86ff167534e)

BibTeX

@misc{dc7df2ff8ff26c,
  title = {Adaptive Localization of Knowledge Negation for Continual LLM Unlearning — arXiv:2611.02620},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.02620},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7df2ff8ff26c; sha256 607f3f23cd371370f95952f3802c5a106a78cdb52e9fcb0ef95cb86ff167534e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.02620. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

With the growing deployment of large language models (LLMs) across diverse domains, concerns regarding their safety have grown substantially. LLM unlearning has emerged as a pivotal approach to removing harmful or unlawful content while maintaining utility. Despite increasing interest, the challenges of continual unlearning, which is common in real-world scenarios, remain underexplored. Successive unlearning tasks often lead to intensified utility degradation. To effectively unlearn targeted knowledge while preserving LLM utility, it is essential to minimize changes in model parameters by selectively updating those linked to the target knowledge, thereby ensuring other knowledge remains unaffected. Building on the task vector framework, we propose a new method named ALKN (Adaptive Localization of Knowledge Negation), which uses dynamic masking to sparsify training gradients and adaptively adjusts unlearning intensity based on inter-task relationships. Comprehensive experiments across three well-established LLM unlearning datasets demonstrate that our approach consistently outperforms baseline methods in both unlearning effectiveness and utility retention under continual unlearning settings.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.02620 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-03T11:02:37.629Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).