arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2004.03761cs.LGcs.AIcs.NE
收藏

RL中的自适应Transformers

Adaptive Transformers in RL

University of Toronto(多伦多大学)

Shakti Kumar, Jerrod Parker, Panteha Naderian

摘要:

Transformers的最新发展为部分可观察强化学习任务的研究开辟了新的有趣领域。2019年底的结果显示,Transformers在记忆密集型和反应型任务上的表现都优于LSTM。在这项工作中,我们首先在反应式和基于内存的环境中部分复制了RL中稳定Transformers的结果。然后,我们表现出性能的改善加上减少计算时,增加自适应注意广度,这个稳定的Transformer上具有挑战性的DMLab30环境。我们所有实验和模型的代码都可以在https://github.com/jerrodparker20/adaptive-transformers-in-rl上找到。

英文摘要:

Recent developments in Transformers have opened new interesting areas of research in partially observable reinforcement learning tasks. Results from late 2019 showed that Transformers are able to outperform LSTMs on both memory intense and reactive tasks. In this work we first partially replicate the results shown in Stabilizing Transformers in RL on both reactive and memory based environments. We then show performance improvement coupled with reduced computation when adding adaptive attention span to this Stable Transformer on a challenging DMLab30 environment. The code for all our experiments and models is available at https://github.com/jerrodparker20/adaptive-transformers-in-rl.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑