Layers with recurrent hidden state

Basically all layers which define get_rec_initial_extra_outputs.

Most (all) of them only when inside a rec loop, otherwise they behave just as if you would iterate over the time axis.

• WindowLayer (optional)
• CumsumLayer
• RecLayer
• RnnCellLayer
• SelfAttentionLayer
• KenLmStateLayer
• CumConcatLayer
• SubnetworkLayer (optional)
• BaseChoiceLayer, ChoiceLayer, etc (internal)
• DecideKeepBeamLayer (internal)
• EditDistanceTableLayer
• MaskedComputationLayer
• UnmaskLayer
• TwoDLSTMLayer