Metraipolis › archive › dc_899edc813cde
Eternal Sunshine of the Spotless Mind: Systematically Erasing LLM's Memories — arXiv:2609.36414
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.36414
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T16:56:31.166Z
- SHA-256 of the archived text
ea1f0c39e557f7371a79e1f315c79679cfd3c8c409d48833db0590d8d252d59a- Stable URL of this page
https://metraipolis.com/documento/dc_899edc813cde
How to cite
Eternal Sunshine of the Spotless Mind: Systematically Erasing LLM's Memories — arXiv:2609.36414 — arXiv (en), https://arxiv.org/abs/2609.36414 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_899edc813cde (Metraipolis archive dc_899edc813cde, sha256 ea1f0c39e557f7371a79e1f315c79679cfd3c8c409d48833db0590d8d252d59a)
BibTeX
@misc{dc899edc813cde,
title = {Eternal Sunshine of the Spotless Mind: Systematically Erasing LLM's Memories — arXiv:2609.36414},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.36414},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_899edc813cde; sha256 ea1f0c39e557f7371a79e1f315c79679cfd3c8c409d48833db0590d8d252d59a},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.36414. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We consider persistent LLMs that accumulate memories of their interactions with a user over time. Such LLMs maintain memories using external storage, which they can query to overcome the limitations of a fixed context window. Such systems have numerous practical applications, as they can draw on all past interactions when responding to user queries. In this paper, we ask whether LLMs can forget information shared with them upon a user's request. We find that current LLMs fail to delete such information---even when they claim to have forgotten it and even when operating with a limited context. To this end, we consider a new direction of study: Deletion of LLM Memories. We show that naively removing messages that match a user's deletion request is insufficient, since conversations naturally introduce message dependencies that cause information to persist. To correctly handle deletion requests, we propose the DeLLM framework. It dynamically constructs relevant context for each LLM query and maintains a provenance graph of messages to determine which ones must be removed during deletion. Our experiments show that DeLLM achieves a high deletion rate while maintaining utility. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.CL). Categorie arXiv: Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2609.36414 — 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 2026-10-07T16:27:03.204Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).