TY  - RPRT
TI  - SPAR: Semantic-Pixel Self-Alignment and Adaptive Routing for Unified Multimodal Models
AU  - Hongxiang Li
AU  - Hongxu Chen
AU  - Chenyang Zhu
AU  - Xiaoshuang Huang
AU  - Jiayin Cai
AU  - Xiaolong Jiang
AU  - Yao Hu
AU  - Long Chen
PY  - 2026
UR  - https://arxiv.org/abs/2606.23041
ID  - 2606.23041
ER  -