Modern neuroscientific techniques that specifically manipulate and measure neuronal activity in behaving animals now allow bridging of the gap from the cellular to the behavioral level. However, in doing so, they also pose new challenges. Research using incompletely defined manipulations in a high-dimensional space without clear hypotheses is likely to suffer from multiple well-known conceptual and statistical problems. In this context it is essential to develop hypotheses with testable implications across levels. Here we propose that a focus on circuit motifs can help achieve this goal. Viewing neural structures as an assembly of circuit motif building blocks is not new. However, recent tool advances have made it possible to extensively map, specifically manipulate, and quantitatively investigate circuit motifs and thereby reexamine their relevance to brain function.

中文翻译：

---

电路模型作为多层次神经科学的概念工具

现代神经科学技术专门操纵和测量行为动物的神经元活动，现在可以弥合从细胞水平到行为水平的差距。然而，在这样做的同时，它们也带来了新的挑战。在没有明确假设的高维空间中使用不完全定义的操作进行研究可能会遇到多个众所周知的概念和统计问题。在这种情况下，开发具有跨层次可检验含义的假设至关重要。在这里，我们建议关注电路图案可以帮助实现这一目标。将神经结构视为电路基元构建块的组合并不是什么新鲜事。然而，最近的工具进步使得广泛地绘制、特别是操纵、