WSVD is a method for efficient low-rank approximation designed to enable fast and efficient execution of Low-Precision Vision-Language Models (VLMs). By applying SVD at a finer granularity (per-head) and using element-wise importance to guide fine-tuning, WSVD achieves significant decoding speedups while maintaining high accuracy.
1@article{wang2026wsvd,
2 title={WSVD: Weighted Low-Rank Approximation for Fast and Efficient Execution of Low-Precision Vision-Language Models},
3 author={Wang, Haiyu and Wang, Yutong and Jiang, Jack and Zhang, Sai Qian},
4 journal={arXiv preprint arXiv:2604.02570},
5 year={2026}
6}
