TY  - RPRT
TI  - Hierarchical Advantage Weighting for Online RL Fine-Tuning of VLAs from Sparse Episode Outcomes
AU  - Tongyan Fang
AU  - Siyuan Huang
AU  - Naiyu Fang
AU  - Ganlong Zhao
AU  - Zhongjin Luo
AU  - Jianbo Liu
AU  - Xiaogang Wang
AU  - Ying Dong
AU  - Hongsheng Li
PY  - 2026
UR  - https://arxiv.org/abs/2606.17043
ID  - 2606.17043
ER  -