Metraipolis › archive › dc_1502db945b2e

Heirloom: Carrying a Phone's Personalization Across Base-Model Generations — arXiv:2611.00650

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.00650
License (as archived)
CC0-1.0
Archived
2026-11-03T11:11:42.380Z
SHA-256 of the archived text
011827f4eb4d9f4f151b55e1bf947b2b6c1307aa75c87ceedc87bcac26f8f4d9
Stable URL of this page
https://metraipolis.com/documento/dc_1502db945b2e

How to cite

Heirloom: Carrying a Phone's Personalization Across Base-Model Generations — arXiv:2611.00650 — arXiv (en), https://arxiv.org/abs/2611.00650 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_1502db945b2e (Metraipolis archive dc_1502db945b2e, sha256 011827f4eb4d9f4f151b55e1bf947b2b6c1307aa75c87ceedc87bcac26f8f4d9)
BibTeX
@misc{dc1502db945b2e,
  title = {Heirloom: Carrying a Phone's Personalization Across Base-Model Generations — arXiv:2611.00650},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00650},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_1502db945b2e; sha256 011827f4eb4d9f4f151b55e1bf947b2b6c1307aa75c87ceedc87bcac26f8f4d9},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00650. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

On-device language models can learn a user's preferences through adapters trained on private data. However, a base-model upgrade may change the layers or vocabulary, leaving the personal adapter incompatible. Replacing it with a non-personal adapter loses much of the personalization, while retraining requires retaining the user's labeled history. We present Heirloom, a system that transfers a personal adapter to a new base model on the phone without retaining labels. Heirloom uses the old personal model as a teacher, collecting its predictions on stored unlabeled inputs, vendor-provided prompts, or prompts generated by the new base. It then unloads the old base and trains an adapter for the new model to match those predictions, keeping only one base model in memory at a time. We evaluate Heirloom across three generation upgrades from two model families, an architecture-preserving refresh, and three personalization tasks. On movie tagging, migration from stored inputs maintains pre-upgrade accuracy and differs from labeled retraining by at most 1.7 percentage points across the three upgrades. On a phone's neural processing unit (NPU), a fixed-length migration for the 4B upgrade takes 38 minutes and consumes 20% of the battery end to end, making the stored-input setting feasible during overnight charging.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2611.00650 — 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 2026-11-03T10:59:52.385Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).