"Layer" is the container of each 3D visualized object (instance) in TensorSpace.
The concept of "layer" in TensorSpace is similar to the meanings in Keras and TensorFlow: plays the role
of basic building blocks and elements (such as convolutional layer, pooling layer,
fully connected layer etc.) of the neural network.
The difference is that the "layer" in TensorSpace does not have any computing capability.
It is a pure container to build the neural network, store and represent data
and features.
The main purpose of TensorSpace "layer" is to be used to 3D visualize the internal layer data and structure.
The API style is similar to the high-level ML APIs, such as Keras and TensorFlow.js.
It is convenient for users to learn and be familiar with TensorSpace quickly.
Fig. 1 - TensorSpace Layers
See Layer for more details.