发布时间：2026-03-01

文章标题：[CVPR 2026] 基于 MoE 的视觉语言大模型量化工作已被CVPR 2026录用

内容：

Chenwei Jia, Baoting Li, Xuchong Zhang*, Mingzhuo Wei, Bochen Lin, Hongbin Sun. Quant Experts: Token-aware Adaptive Error Reconstruction with Mixture of Experts for Large Vision-Language Models Quantization. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2026.