A Consciousness-Inspired Planning Agent for Model-Based Reinforcement Learning
Abstract
We present an end-to-end, model-based deep reinforcement learning agent which dynamically attends to relevant parts of its state during planning. The agent uses a bottleneck mechanism over a set-based representation to force the number of entities to which the agent attends at each planning step to be small. In experiments, we investigate the bottleneck mechanism with several sets of customized environments featuring different challenges. We consistently observe that the design allows the planning agents to generalize their learned task-solving abilities in compatible unseen environments by attending to the relevant objects, leading to better out-of-distribution generalization performance.
Cite
Text
Zhao et al. "A Consciousness-Inspired Planning Agent for Model-Based Reinforcement Learning." Neural Information Processing Systems, 2021.Markdown
[Zhao et al. "A Consciousness-Inspired Planning Agent for Model-Based Reinforcement Learning." Neural Information Processing Systems, 2021.](https://mlanthology.org/neurips/2021/zhao2021neurips-consciousnessinspired/)BibTeX
@inproceedings{zhao2021neurips-consciousnessinspired,
title = {{A Consciousness-Inspired Planning Agent for Model-Based Reinforcement Learning}},
author = {Zhao, Mingde and Liu, Zhen and Luan, Sitao and Zhang, Shuyuan and Precup, Doina and Bengio, Yoshua},
booktitle = {Neural Information Processing Systems},
year = {2021},
url = {https://mlanthology.org/neurips/2021/zhao2021neurips-consciousnessinspired/}
}