Metraipolis › archive › dc_0272cd7fdb17
Does Personalization Break Trust? Open-Set Rejection Calibration Under Few-Shot On-Device Intent Adaptation — arXiv:2610.20652
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.20652
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T09:52:05.401Z
- SHA-256 of the archived text
4ed413a55968e64ce008d5508739e1238a08d2ca62d0eea0b8f767f476f42b99- Stable URL of this page
https://metraipolis.com/documento/dc_0272cd7fdb17
How to cite
Does Personalization Break Trust? Open-Set Rejection Calibration Under Few-Shot On-Device Intent Adaptation — arXiv:2610.20652 — arXiv (en), https://arxiv.org/abs/2610.20652 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0272cd7fdb17 (Metraipolis archive dc_0272cd7fdb17, sha256 4ed413a55968e64ce008d5508739e1238a08d2ca62d0eea0b8f767f476f42b99)
BibTeX
@misc{dc0272cd7fdb17,
title = {Does Personalization Break Trust? Open-Set Rejection Calibration Under Few-Shot On-Device Intent Adaptation — arXiv:2610.20652},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.20652},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0272cd7fdb17; sha256 4ed413a55968e64ce008d5508739e1238a08d2ca62d0eea0b8f767f476f42b99},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.20652. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Always-on virtual assistants increasingly personalize on-device to new user-specific commands via few-shot fine-tuning (e.g., LoRA), but this adaptation is evaluated almost exclusively on whether it learns the new commands, not on whether the assistant can still recognize an input that matches none of its intents - the out-of-scope (OOS) rejection problem. We show that naive few-shot LoRA personalization degrades OOS-rejection AUROC (area under the ROC curve) on every one of five public intent-classification benchmarks (mean drop $0.055$--$0.258$, five seeds each), while in-scope accuracy is preserved. A scorer-invariant check shows part of this drop is an artifact of the standard confidence scorer's own miscalibration under adaptation, though a smaller, genuine effect remains under any scorer; we trace it to a measurable shrinkage of the embedding margin around known-intent regions (pooled Spearman rho=0.61$). A deployment-cheap fix is scoring rejection by Mahalanobis distance instead of softmax confidence without extra training/retraining recovers most of the lost quality on 3 of 5 datasets at no accuracy cost, beats two further OOD scoring baselines, and generalizes across shot count, LoRA rank, and a second base encoder. We report honestly where the fix falls short and release all code for reproducibility. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno). Categorie arXiv: Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2610.20652 — 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 2026-10-16T08:44:47.605Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).