theanets.layers.recurrent.ARRNN

class theanets.layers.recurrent.ARRNN(size, inputs, name=None, activation='relu', **kwargs)

An RNN with adaptive rate per unit.

Notes

In a normal RNN, a hidden unit is updated completely at each time step, \(h_t = f(x_t, h_{t-1})\). With an explicit update rate, the state of a hidden unit is computed as a mixture of the new and old values,

\[h_t = (1 - z_t) \odot h_{t-1} + z_t \odot f(x_t, h_{t-1})\]

where \(\odot\) indicates elementwise multiplication.

Rates might be defined in a number of ways, spanning a continuum between vanilla RNNs (i.e., all rate parameters are fixed at 1) all the way to parametric rates that are computed as a function of the inputs and the hidden state at each time step (i.e., something more like the GRU).

In the ARRNN model, the rate values are represented as a computed at each time step as a logistic sigmoid applied to an affine transform of the input:

\[z_t = \frac{1}{1 + \exp(-x_t W_{xr} - b_r)}.\]

This representation of the rates uses more parameters than the LRRNN but is able to adapt rates to the input at each time step. However, in this model, rates are not able to adapt to the state of the hidden units at each time step.

Parameters

• b — vector of bias values for each hidden unit
• r — vector of rate biases for each hidden unit
• xh — matrix connecting inputs to hidden units
• xr — matrix connecting inputs to rate “gates”
• hh — matrix connecting hiddens to hiddens

Outputs

• out — the post-activation state of the layer
• pre — the pre-activation state of the layer
• hid — the pre-rate-mixing hidden state
• rate — the rate values

__init__(size, inputs, name=None, activation='relu', **kwargs)

Methods

add_weights(name, nin, nout[, mean, std, ...])	Helper method to create a new weight matrix.
initial_state(name, batch_size)	Return an array of suitable for representing initial state.
setup()	Set up the parameters and initial values for this layer.
transform(inputs)	Transform the inputs for this layer into an output for the layer.

Attributes

input_size	For networks with one input, get the input size.
num_params	Total number of learnable parameters in this layer.
params	A list of all parameters in this layer.

setup()

Set up the parameters and initial values for this layer.

transform(inputs)

Transform the inputs for this layer into an output for the layer.

Parameters:

inputs : dict of theano expressions

Symbolic inputs to this layer, given as a dictionary mapping string names to Theano expressions. See base.Layer.connect().

Returns:

outputs : theano expression

A map from string output names to Theano expressions for the outputs from this layer. This layer type generates a “pre” output that gives the unit activity before applying the layer’s activation function, a “hid” output that gives the rate-independent, post-activation hidden state, a “rate” output that gives the rate value for each hidden unit, and an “out” output that gives the hidden output.

updates : list of update pairs

A sequence of updates to apply inside a theano function.