Layers with special behavior on dynamic spatial axes

This is a list of layers with special behavior on dynamic spatial axes, i.e. axes with dynamic sequence lengths because considering the padding or sequence lengths is important for correct behavior.

• SoftmaxOverSpatialLayer, will make sure that the padded frames are masked away.
• BatchSoftmaxLayer
• ReduceLayer. sum, max etc will ignore the padded frames.
• MathNormLayer. shares code with ReduceLayer internally.
• BatchNormLayer (and batch norm in general on any layer)
• (NormLayer actually should have special behavior and ignore padded frames, but incorrectly it does not currently)
• SliceNdLayer
• SeqLenMaskLayer
• FlattenBatchLayer
• PostfixInTimeLayer
• (CumsumLayer with reverse=True should ignore padded frames but currently does not (#574))

(This list is currently incomplete.)

Somewhat related is the option recurrent on each layer class. recurrent=False implies that sequence lengths do not matter as well as the ordering of frames. But this is not exactly the same. E.g. ConvLayer has recurrent=True but ConvLayer does not make use of the sequence lengths.

The obvious example of a layer where the dynamic spatial axes do not matter is LinearLayer.