On-device large language models: a survey of model compression and system optimization
Authors:
- Wanyi ChenWCWanyi Chen
- Junhao WangJWJunhao Wang
- Yiwei ZhangYZYiwei Zhang
- Yufan ShiYSYufan Shi
- Tianyi JiangTJTianyi Jiang
- Shengxian ZhouSZShengxian Zhou
- Chenxu WuCWChenxu Wu
- Andi ZhangAZAndi Zhang
- Chenyue ZhouCZChenyue Zhou
- Minxuan WangMWMinxuan Wang
- Xinyu LiuXLXinyu Liu
- Xiaoshuai HaoXHXiaoshuai Hao
Open Access
- 22 May 2026
- journal article
- Published by Springer Nature in Artificial Intelligence Review
- Vol. 59 (9)
Abstract
No abstract availableKeywords
All Related Versions
This publication has 128 references indexed in Scilit:
- Z -Prunear: Post-Training Pruning of Large Language Models for Efficiency without Retraining2025
- Enhancing Video-LLM Reasoning via Agent-of-Thoughts Distillation2025
- Align-KD: Distilling Cross-Modal Alignment Knowledge for Mobile Vision-Language Large Model Enhancement2025
- Optimising TinyML with quantization and distillation of transformer and mamba models for indoor localisation on edge devicesScientific Reports, 2025
- Markov Knowledge Distillation: Make Nasty Teachers Trained by Self-undermining Knowledge Distillation Fully Distillable2024
- Robust knowledge distillation based on feature variance against backdoored teacher modelApplied Soft Computing, 2024
- PromptKD: Unsupervised Prompt Distillation for Vision-Language Models2024
- Empowering generative AI through mobile edge computingNature Reviews Electrical Engineering, 2024
- Loss aware post-training quantizationMachine Learning, 2021
- Benchmarking Contemporary Deep Learning Hardware and Frameworks: A Survey of Qualitative Metrics2019