An Overview of Low-Rank Structures in the Training and Adaptation of Large Models

The rise of deep learning has revolutionized data processing and prediction in signal processing and machine learning, yet the substantial computational demands of training and deploying modern large-scale deep models present significant challenges, including high computational costs and energy consumption. Recent research has uncovered a widespread phenomenon in deep networks: the emergence of low-rank structures in weight matrices and learned representations during training. These implicit low-dimensional patterns provide valuable insights for improving the efficiency of training and fine-tuning large-scale models. Practical techniques inspired by this phenomenon, such as low-rank adaptation (LoRA) and training, enable significant reductions in computational cost while preserving model performance. In this paper, we present a comprehensive review of recent advances in exploiting low-rank structures for deep learning and shed light on their mathematical foundations. Mathematically, we present two complementary perspectives on understanding the low-rankness in deep networks: (i) the emergence of low-rank structures throughout the whole optimization dynamics of gradient and (ii) the implicit regularization effects that induce such low-rank structures at convergence. From a practical standpoint, studying the low-rank learning dynamics of gradient descent offers a mathematical foundation for understanding the effectiveness of LoRA in fine-tuning large-scale models and inspires parameter-efficient low-rank training strategies. Furthermore, the implicit low-rank regularization effect helps explain the success of various masked training approaches in deep neural networks, ranging from dropout to masked self-supervised learning.

@article{balzano2025_2503.19859,
  title={ An Overview of Low-Rank Structures in the Training and Adaptation of Large Models },
  author={ Laura Balzano and Tianjiao Ding and Benjamin D. Haeffele and Soo Min Kwon and Qing Qu and Peng Wang and Zhangyang Wang and Can Yaras },
  journal={arXiv preprint arXiv:2503.19859},
  year={ 2025 }
}