paperJune 2026Unreviewed

Forget to Improve: On-Device LLM-Agent Continual Learning via Budget-Curated Memory

Beining Wu, Zihao Ding, Jun Huang, Yanxiao Zhao

Abstract

On-device language-model agents improve by accumulating experience in retrieved memory rather than by updating weights. This memory is hard-bounded and exposed: it consumes RAM and energy, reaches peers through a thin uplink, and becomes an attack surface because it is writable by what the agent reads. Existing systems each cover one part of this problem: agentic memories grow without a budget, on-device methods keep entries by success alone, and poisoning is studied mainly as an attack rather t

Framework mappings

OWASP Top 10 for LLM Applications

LLM04Data and Model Poisoning

MITRE ATLAS

AML.T0020Poison Training Data

Suggested from the entry's categories.

Cite

@misc{wu2026forget,
  title = {{Forget to Improve: On-Device LLM-Agent Continual Learning via Budget-Curated Memory}},
  author = {Beining Wu and Zihao Ding and Jun Huang and Yanxiao Zhao},
  year = {2026},
  month = jun,
  eprint = {2606.25115},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2606.25115}
}