AI Bibliography |
Lechner, M., Hasani, R. M., & Grosu, R. (2018). Neuronal circuit policies. arXiv preprint arXiv:1803.08554. |
Resource type: Journal Article BibTeX citation key: Lechner2018 View all bibliographic details |
Categories: Artificial Intelligence, Biological Science, Computer Science, Data Sciences, Decision Theory, General, Neuroscience Subcategories: Autonomous systems, Big data, Decision making, Deep learning, Machine learning, Neural nets, Neurosymbolic Creators: Grosu, Hasani, Lechner Publisher: Collection: arXiv preprint arXiv:1803.08554 |
Attachments |
Abstract |
We propose an effective way to create inter- pretable control agents, by re-purposing the func- tion of a biological neural circuit model, to gov- ern simulated and real world reinforcement learning (RL) test-beds. We model the tap-withdrawal (TW) neural circuit of the nematode, C. elegans, a circuit responsible for the worm’s reflexive response to external mechanical touch stimulations, and learn its synaptic and neuronal parameters as a policy for controlling basic RL tasks. We also autonomously park a real rover robot on a pre- defined trajectory, by deploying such neuronal circuit policies learned in a simulated environment. For reconfiguration of the purpose of the TW neural circuit, we adopt a search-based RL algorithm. We show that our neuronal policies perform as good as deep neural network policies with the advantage of realizing interpretable dynamics at the cell level. |