TY  - RPRT
TI  - DenseGRPO: From Sparse to Dense Reward for Flow Matching Model Alignment
AU  - Haoyou Deng
AU  - Keyu Yan
AU  - Chaojie Mao
AU  - Xiang Wang
AU  - Yu Liu
AU  - Changxin Gao
AU  - Nong Sang
PY  - 2026
UR  - https://arxiv.org/abs/2601.20218
ID  - 2601.20218
ER  -