AAAI · 2026 · Conference paper

Put the Space of LoRA Initialization to the Extreme to Preserve Pre-trained Knowledge

Pengwei Tang, Xiaolin Hu, Yong Liu, Lizhong Ding, Dongjie Zhang, Xing Wu, Debing Zhang

Beijing Academy of Artificial Intelligence · Ministry of Education · Beijing Institute of Technology · Xiaomi (China) · Institute of Information Engineering

Read here PDF ↗

Rendered by arXiv from the LaTeX source. If anything looks wrong, switch to the PDF.HTML · PDF · Open in a new tab ↗ · Close

Published in: Proceedings of the AAAI Conference on Artificial Intelligence
Date: 2026-03-14
Citations: 12
arXiv: 2503.02659 ↗

Cite

@inproceedings{tang2026space,
  title = {Put the Space of LoRA Initialization to the Extreme to Preserve Pre-trained Knowledge},
  author = {Pengwei Tang and Xiaolin Hu and Yong Liu and Lizhong Ding and Dongjie Zhang and Xing Wu and Debing Zhang},
  year = {2026},
  booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence},
  eprint = {2503.02659},
  archivePrefix = {arXiv},
  doi = {10.1609/aaai.v40i39.40608},
  url = {https://arxiv.org/abs/2503.02659},
}