notesum.ai
Published at October 22Corrected Soft Actor Critic for Continuous Control
cs.LG
cs.AI
Released Date: October 22, 2024
Authors: Yanjun Chen1, Xinming Zhang2, Xianghui Wang2, Zhiqiang Xu2, Xiaoyu Shen2, Wei Zhang2
Aff.: 1Department of Computing, The Hong Kong Polytechnic University; 2Digital Twin Institute, Eastern Institute of Technology, Ningbo, China

| Parameter | Value |
|---|---|
| Temperature Parameter () | 0.2 |
| Neural Network Architecture | [256, 256] |
| Learning Rate | |
| Discount Factor () | 0.99 |
| Soft Update Coefficient () | 0.005 |
| Initial Exploration Steps | 10,000 |
| Training Episodes | 1,000 |
| Number of Seeds | 10 |