TY  - RPRT
TI  - Forge-and-Quench: Enhancing Image Generation for Higher Fidelity in Unified Multimodal Models
AU  - Yanbing Zeng
AU  - Jia Wang
AU  - Hanghang Ma
AU  - Junqiang Wu
AU  - Jie Zhu
AU  - Xiaoming Wei
AU  - Jie Hu
PY  - 2026
UR  - https://arxiv.org/abs/2601.04706
ID  - 2601.04706
ER  -