稀疏结构能显著提升图像强化学习性能，效果取决于权重是否可训练。
On the Interplay Between Sparsity and Training in Deep Reinforcement Learning
- 对比多种同容量稀疏结构，发现其对学习效果影响显著。
- 固定权重时，空间偏置结构表现更优；可学习时，全连接结构更佳。
- 为不同场景选择合适稀疏结构，可提升强化学习效率。
我们研究了深度强化学习中不同稀疏架构的收益。重点关注图像域，其中空间偏置和全连接架构较为常见。在与多种同容量架构对比后，发现稀疏结构对学习性能有显著影响。此外，针对特定任务选择最优稀疏结构，取决于隐藏层权重是否可学习：固定权重时，空间偏置架构表现更优；若权重可学习，则全连接架构更具优势。
原文摘要 · Abstract (English)
We study the benefits of different sparse architectures for deep reinforcement learning. In particular, we focus on image-based domains where spatially-biased and fully-connected architectures are common. Using these and several other architectures of equal capacity, we show that sparse structure has a significant effect on learning performance. We also observe that choosing the best sparse architecture for a given domain depends on whether the hidden layer weights are fixed or learned.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。