TY  - RPRT
TI  - Hybrid-Adaptive Thread Tuning to Mitigate Simulation Execution Bottlenecks in High-Performance Reinforcement Learning Inference
AU  - Jiming Su
AU  - Hantao Hua
AU  - Lujia Yin
AU  - Yiping Yao
AU  - Feng Zhu
PY  - 2026
UR  - https://arxiv.org/abs/2608.06025
ID  - 2608.06025
ER  -