TY  - RPRT
TI  - DR-APO: Decoupled-Representation Advantage-Guided Policy Optimization from Continuous Control to Embodied Navigation
AU  - Runhua Zhang
AU  - Yingjie Niu
AU  - Chao Li
PY  - 2026
UR  - https://arxiv.org/abs/2610.16618
ID  - 2610.16618
ER  -