Priority-Aware Learning-Unlearning Correction for Dynamic Decentralized LoRA Fine-Tuning
面向动态去中心化LoRA微调的优先级感知学习-遗忘校正
机构 * Beijing Key Laboratory of Network System Architecture and Convergence, Beijing University of Posts and Telecommunications(网络系统架构与融合北京市重点实验室,北京邮电大学) ; Information Systems Technology and Design Pillar, Singapore University of Technology and Design(新加坡科技设计大学信息系统技术与设计系)
专题命中 指令微调 :large language model(abstract);language model(abstract);post-training(abstract);分类 cs.AI、cs.LG
AI总结 针对去中心化边缘网络中设备动态加入/离开导致的参数纠缠问题,提出基于正交LoRA的优先级感知校正框架,通过拓扑调整和资源分配实现高效知识评估与修正。
Comments 15 pages, 8 figures