TY  - RPRT
TI  - Infinite Sampling: Efficient and Stable Grouped RL Training for Large Language Models
AU  - Liangyu Wang
AU  - Huanyi Xie
AU  - Xinhai Wang
AU  - Tianjin Huang
AU  - Mengdi Li
AU  - Di Wang
PY  - 2025
UR  - https://arxiv.org/abs/2506.22950
ID  - 2506.22950
ER  -