Metraipolis › archive › dc_bc1421837cbd

Learning Personalized Memory Gating from Delayed One-Sided Feedback — arXiv:2610.23812

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.23812
License (as archived)
CC0-1.0
Archived
2026-10-19T09:10:28.862Z
SHA-256 of the archived text
553745428e791d8c0d45dbb882818abb9b2ef10660d1674ee8838b832d379da8
Stable URL of this page
https://metraipolis.com/documento/dc_bc1421837cbd

How to cite

Learning Personalized Memory Gating from Delayed One-Sided Feedback — arXiv:2610.23812 — arXiv (en), https://arxiv.org/abs/2610.23812 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bc1421837cbd (Metraipolis archive dc_bc1421837cbd, sha256 553745428e791d8c0d45dbb882818abb9b2ef10660d1674ee8838b832d379da8)
BibTeX
@misc{dcbc1421837cbd,
  title = {Learning Personalized Memory Gating from Delayed One-Sided Feedback — arXiv:2610.23812},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.23812},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bc1421837cbd; sha256 553745428e791d8c0d45dbb882818abb9b2ef10660d1674ee8838b832d379da8},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.23812. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Long-term memory enables language-model agents to personalize future interactions, but a finite memory bank cannot preserve every extracted fact. Existing memory gating policies are typically shared across users and therefore fail to account for user-specific patterns of future reuse. In this paper, we study memory gating at the fact level. We first propose a retrieval-based labeling strategy that labels each fact by whether it is later retrieved and incorporated into responses. We then formulate online memory gating as contextual apple tasting with delayed one-sided feedback, and propose Warm-Start UCB, which combines offline learning from existing data with online adaptation to the current user. Our theoretical analysis establishes a high-probability regret upper bound of $\tilde{O}(T\epsilon \wedge d\sqrt{T})$ when the delay window is at most $O(\sqrt{T})$, where $T$ is the total number of rounds, $\epsilon$ is the initialization error, $d$ is the dimension of the feature map, and $\tilde{O}(\cdot)$ hides constant and poly-logarithmic factors. The results of our experiment demonstrate that our method outperforms all baselines under different memory budgets and history lengths, establishing it as a promising personalized memory gating policy.

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/2610.23812 — 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 2026-10-19T09:06:29.574Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).