arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2201.05242cs.LGcs.NEcs.ROq-bio.NC
收藏

用于并行控制的神经电路结构先验

Neural Circuit Architectural Priors for Embodied Control

Cold Spring Harbor Laboratory(冷泉港实验室)

Nikhil X. Bhattasali, Anthony M. Zador, Tatiana A. Engel

摘要:

用于电机控制的人工神经网络通常采用通用架构,如完全连接的MLP。虽然一般来说,这些白板架构依赖于大量的经验来学习,不容易转移到新的机构,并且具有难以解释的内部动态。在自然界中,动物天生就具有高度结构化的神经系统连接,这些神经系统是由进化形成的;这种先天回路与学习机制协同作用,提供诱导偏差,使大多数动物在出生后不久就能正常工作并有效学习。受视觉电路启发的卷积网络已经编码了视觉的有用偏差。然而,目前还不清楚受神经电路启发的ANN架构在多大程度上可以为其他AI领域产生有用的偏差。在这项工作中,我们问生物启发的人工神经网络架构可以提供在电机控制领域的优势。具体来说,我们翻译C。elegans运动电路到人工神经网络模型控制模拟游泳代理。在运动任务上,我们的架构实现了良好的初始性能和渐近性能与MLP相当,同时大大提高了数据效率,并需要数量级更少的参数。我们的架构是可解释的,并转移到新的车身设计。消融分析表明,约束的激励/抑制是至关重要的学习,而权重初始化有助于良好的初始性能。我们的工作证明了生物启发的人工神经网络架构的几个优点,并鼓励未来的工作在更复杂的体现控制。

英文摘要:

Artificial neural networks for motor control usually adopt generic architectures like fully connected MLPs. While general, these tabula rasa architectures rely on large amounts of experience to learn, are not easily transferable to new bodies, and have internal dynamics that are difficult to interpret. In nature, animals are born with highly structured connectivity in their nervous systems shaped by evolution; this innate circuitry acts synergistically with learning mechanisms to provide inductive biases that enable most animals to function well soon after birth and learn efficiently. Convolutional networks inspired by visual circuitry have encoded useful biases for vision. However, it is unknown the extent to which ANN architectures inspired by neural circuitry can yield useful biases for other AI domains. In this work, we ask what advantages biologically inspired ANN architecture can provide in the domain of motor control. Specifically, we translate C. elegans locomotion circuits into an ANN model controlling a simulated Swimmer agent. On a locomotion task, our architecture achieves good initial performance and asymptotic performance comparable with MLPs, while dramatically improving data efficiency and requiring orders of magnitude fewer parameters. Our architecture is interpretable and transfers to new body designs. An ablation analysis shows that constrained excitation/inhibition is crucial for learning, while weight initialization contributes to good initial performance. Our work demonstrates several advantages of biologically inspired ANN architecture and encourages future work in more complex embodied control.

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。

↑