GRU Projected Layer

Gated recurrent unit (GRU) projected layer for recurrent neural network (RNN)

Since R2025a

• 

Libraries:
Deep Learning Toolbox / Deep Learning Layers / Sequence Layers

Description

The GRU Projected Layer block represents a recurrent neural network (RNN) layer that learns dependencies between time steps in time-series and sequence data in the CT format (two dimensions corresponding to channels and time steps, in that order) by using projected learnable weights.

To compress a deep learning network, you can use projected layers. A projected layer is a type of deep learning layer that enables compression by reducing the number of stored learnable parameters. The layer introduces learnable projector matrices Q, replaces multiplications of the form $$Wx$$, where W is a learnable matrix, with the multiplication $$WQ{Q}^{\top }x$$, and stores Q and $$W\prime =WQ$$ instead of storing W. Projecting x into a lower dimensional space using Q typically requires less memory to store the learnable parameters and can have similarly strong prediction accuracy.

Reducing the number of learnable parameters by projecting a GRU layer rather than reducing the number of hidden units of the GRU layer maintains the output size of the layer and, in turn, the sizes of the downstream layers, which can result in better prediction accuracy.

The exportNetworkToSimulink function generates this block to represent a gruProjectedLayer object.

Limitations

• The Layer parameter does not accept gruProjectedLayer objects that have the HasStateInputs or HasStateOutputs properties set to 1 (true).

Ports

Input

Port_1 — Input data

vector | 2-dimensional array

Input data. The data must have two dimensions corresponding to channels and time steps, in that order, or one dimension corresponding to channels.

Data Types: single | double | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | uint64 | fixed point

Output

Port_1 — Output data

2-dimensional array

The result of applying the GRU projected operation to the input data. The output data has two dimensions corresponding to channels and time steps, in that order.

Data Types: single | double | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | uint64 | fixed point

Parameters

To edit block parameters interactively, use the Property Inspector. From the Simulink^® Toolstrip, on the Simulation tab, in the Prepare gallery, select Property Inspector.

Main

Layer — Corresponding layer object

layerObject (default) | gruProjectedLayer object

Specify the name of a workspace variable that contains a gruProjectedLayer object from a trained network. The GRU Projected Layer block configures itself by using the properties of the object and calculates the block output by using the learnable parameters of the object.

Programmatic Use

Block Parameter: Layer

Type: workspace variable

Values: gruProjectedLayer object

Default: 'layerObject'

Data format — Format of input data

CT (default)

Data format for the input data. The options use the same notation as the fmt argument of the dlarray object, except layer blocks do not support the Batch (B) dimension and instead assume an observation number of 1.

Programmatic Use

Block Parameter: DataFormat

Type: character vector

Values: 'CT'

Default: 'CT'

Stateful prediction — Whether to use stateful prediction

on (default) | off

Whether to use stateful prediction, specified as a boolean. If true, the block maintains the hidden state between time steps. If false, the block performs stateless prediction by resetting the hidden states to their initial values at the beginning of each time step. Use stateless prediction for frame-based processing where Simulink step time represents frame period and the network processes multiple samples at each time step. For more information, see Sample- and Frame-Based Concepts (DSP System Toolbox).

Programmatic Use

Block Parameter: StatefulPrediction

Type: character vector

Values: 'on' | 'off'

Default: 'on'

Data Types

Activation approximation method — Approximation method for computing activation function output

None (default) | CORDIC | Lookup

If the object that you pass as the value of the Layer parameter uses the tanh state activation function or the sigmoid gate activation function, then the block uses the approximation method that you specify to compute the layer output.

Approximation Method	Data Types Supported	When to Use This Method
None (default)	Floating-point	You are processing only floating-point data.
CORDIC	Floating-point (double and single) and fixed-point with a Bias value of 0 and a Slope value of a power of 2	You are processing fixed-point data and want to deploy to FPGA hardware.
Lookup	Floating-point and fixed-point	You are processing fixed-point data and want to generate C/C++ code.

For more information about the CORDIC approximation method, see cordictanh (Fixed-Point Designer).

Programmatic Use

Block Parameter: ApproximationMethod

Type: character vector

Values: 'None' | 'CORDIC' | 'Lookup'

Default: 'None'

Output minimum — Minimum output value for range checking

[] (default) | scalar

Lower value of the output range that the software checks.

The software uses the minimum to perform:

• Parameter range checking (see Specify Minimum and Maximum Values for Block Parameters (Simulink)) for some blocks.

• Simulation range checking (see Specify Signal Ranges (Simulink) and Enable Simulation Range Checking (Simulink)).

• Automatic scaling of fixed-point data types.

• Optimization of the code that you generate from the model. This optimization can remove algorithmic code and affect the results of some simulation modes such as SIL or external mode. For more information, see Optimize using the specified minimum and maximum values (Embedded Coder).

Tips

Output minimum does not saturate or clip the actual output signal. Use the Saturation (Simulink) block instead.

Dependencies

To enable this parameter, set Output data type to a value other than Inherit: Inherit via internal rule.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	OutMin
Values:	'[]' (default) | scalar in quotes

Output maximum — Maximum output value for range checking

[] (default) | scalar

Upper value of the output range that the software checks.

The software uses the maximum value to perform:

• Parameter range checking (see Specify Minimum and Maximum Values for Block Parameters (Simulink)) for some blocks.

• Simulation range checking (see Specify Signal Ranges (Simulink) and Enable Simulation Range Checking (Simulink)).

• Automatic scaling of fixed-point data types.

• Optimization of the code that you generate from the model. This optimization can remove algorithmic code and affect the results of some simulation modes such as SIL or external mode. For more information, see Optimize using the specified minimum and maximum values (Embedded Coder).

Tips

Output maximum does not saturate or clip the actual output signal. Use the Saturation (Simulink) block instead.

Dependencies

To enable this parameter, set Output data type to a value other than Inherit: Inherit via internal rule.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	OutMax
Values:	'[]' (default) | scalar in quotes

Output data type — Data type of output

Inherit: Inherit via internal rule (default) | <data type expression>

Choose the data type for the output. The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType. When you choose Inherit: Inherit via internal rule, Simulink sets the output data type to the same type as the hidden state.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	OutDataTypeStr
Values:	'Inherit: Inherit via internal rule' (default) | <data type expression>

Lock output data type setting against changes by the fixed-point tools — Option to prevent fixed-point tools from overriding Output data type

off (default) | on

Select this parameter to prevent the fixed-point tools from overriding the Output data type you specify on the block. For more information, see Use Lock Output Data Type Setting (Fixed-Point Designer).

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	LockScale
Values:	'off' (default) | 'on'

Integer rounding mode — Rounding mode for fixed-point operations

Floor (default) | Ceiling | Convergent | Nearest | Round | Simplest | Zero

Specify the rounding mode for fixed-point operations. For more information, see Rounding Modes (Fixed-Point Designer).

Block parameters always round to the nearest representable value. To control the rounding of a block parameter, enter an expression using a MATLAB^® rounding function in the mask field.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	RndMeth
Values:	'Floor' (default) | 'Ceiling' | 'Convergent' | 'Nearest' | 'Round' | 'Simplest' | 'Zero'

Saturate on integer overflow — Method of overflow action

off (default) | on

Specify whether integer overflows saturate or wrap.

• on — Overflows saturate to either the minimum or maximum value that the data type can represent.

• off — Overflows wrap to the appropriate value that the data type can represent.

For example, the maximum value that the signed 8-bit integer int8 can represent is 127. Any block operation result greater than the maximum value causes overflow of the 8-bit integer.

• With this parameter selected, the block output saturates at 127. Similarly, the block output saturates at a minimum output value of –128.

• With this parameter cleared, the software interprets the overflow-causing value as int8, which can produce an unintended result. For example, a block result of 130 (binary 1000 0010) expressed as int8 is –126.

Tips

• Set this parameter to on when your model has a possible overflow and you want explicit saturation protection in the generated code.

• To optimize the efficiency of your generated code, keep the default off setting for this parameter. Using the default setting also helps you avoid overspecifying how the block handles out-of-range signals. For more information, see Troubleshoot Signal Range Errors (Simulink).

• When you select this parameter, saturation applies to every internal operation on the block, not just the output or the result.

• In general, the code generation process can detect when overflow is not possible. In this case, the code generator does not produce saturation code.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	SaturateOnIntegerOverflow
Values:	'off' (default) | 'on'

Input weights — Data type of input weights

Inherit: Inherit via back propagation (default) | Inherit: Inherit from 'Constant value' | <data type expression>

The block casts the value of the InputWeights property of the object that you specify with the Layer parameter to this data type. The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	InputWeightsDataTypeStr
Values:	'Inherit: Inherit via back propagation' (default) | 'Inherit: Inherit from 'Constant value'' | <data type expression>

Recurrent weights — Data type of recurrent weights

Inherit: Inherit via back propagation (default) | Inherit: Inherit from 'Constant value' | <data type expression>

The block casts the value of the RecurrentWeights property of the object that you specify with the Layer parameter to this data type. The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	RecurrentWeightsDataTypeStr
Values:	'Inherit: Inherit via back propagation' (default) | 'Inherit: Inherit from 'Constant value'' | <data type expression>

Bias — Data type of bias

Inherit: Inherit via back propagation (default) | Inherit: Inherit from 'Constant value' | <data type expression>

The block casts the value of the Bias property of the object that you specify with the Layer parameter to this data type. The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	BiasDataTypeStr
Values:	'Inherit: Inherit via back propagation' (default) | 'Inherit: Inherit from 'Constant value'' | <data type expression>

Input projector weights — Data type of input projector weights

Inherit: Inherit via back propagation (default) | Inherit: Inherit from 'Constant value' | <data type expression>

The block casts the value of the InputProjector property of the object that you specify with the Layer parameter to this data type. The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType.

For more information, see GRU Projected Layer.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	InputProjectorWeightsDataTypeStr
Values:	'Inherit: Inherit via back propagation' (default) | 'Inherit: Inherit from 'Constant value'' | <data type expression>

Output projector weights — Data type of output projector weights

Inherit: Inherit via back propagation (default) | Inherit: Inherit from 'Constant value' | <data type expression>

The block casts the value of the OutputProjector property of the object that you specify with the Layer parameter to this data type. The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType.

For more information, see GRU Projected Layer.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	OutputProjectorWeightsDataTypeStr
Values:	'Inherit: Inherit via back propagation' (default) | 'Inherit: Inherit from 'Constant value'' | <data type expression>

Initial hidden state — Data type of initial hidden state

Inherit: Inherit via back propagation (default) | Inherit: Inherit from 'Constant value' | <data type expression>

The block casts the value of the HiddenState property of the object that you specify with the Layer parameter to this data type. The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType.

This parameter affects only the initial hidden state, h₀. To cast later hidden state values, use the Hidden state parameter. For more information, see Gated Recurrent Unit Layer.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	InitialHiddenStateDataTypeStr
Values:	'Inherit: Inherit via back propagation' (default) | 'Inherit: Inherit from 'Constant value'' | <data type expression>

Hidden state — Data type of hidden state

Inherit: Inherit via internal rule (default) | Inherit: Keep MSB | Inherit: Match scaling | Inherit: Inherit via back propagation | Inherit: Same as first input | <data type expression>

Choose the data type for the output of the subsystem ForIteratorSubsystem/HiddenState inside the GRU Projected Layer block. The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType. When you select Inherit: Inherit via internal rule, Simulink chooses a data type to balance numerical accuracy, performance, and generated code size, while taking into account the properties of the embedded target hardware.

For a time step t, the subsystem computes the hidden state h_t as

$$h_t=(1-z_t)\odot {\tilde{h}}_t+z_t\odot h_{t-1}.$$

, where ⊙ denotes the Hadamard product (element-wise multiplication of vectors). For more information, see Gated Recurrent Unit Layer.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	HiddenStateDataTypeStr
Values:	'Inherit: Inherit via internal rule' (default) | 'Inherit: Keep MSB' | 'Inherit: Keep LSB' | 'Inherit: Inherit via back propagation' | 'Inherit: Same as first input' | 'Inherit: Same as accumulator' | <data type expression>

Input weights matrix multiply output — Data type of input weights matrix multiply output

Inherit: Inherit via internal rule (default) | Inherit: Keep MSB | Inherit: Match scaling | Inherit: Inherit via back propagation | Inherit: Same as first input | <data type expression>

Choose the data type for the output of the Product block InputWeightsMatrixMultiply/W*x inside the GRU Projected Layer block. The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType. When you select Inherit: Inherit via internal rule, Simulink chooses a data type to balance numerical accuracy, performance, and generated code size, while taking into account the properties of the embedded target hardware.

For a time step t, the Product block computes the product of the input weights (W) and the input at the time step (x_t). For more information, see Gated Recurrent Unit Layer.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	InputWeightsMatrixMulitplyOutDataTypeStr
Values:	'Inherit: Inherit via internal rule' (default) | 'Inherit: Keep MSB' | 'Inherit: Match scaling' | 'Inherit: Inherit via back propagation' | 'Inherit: Same as first input' | <data type expression>

Recurrent weights matrix multiply output — Data type of recurrent weights matrix multiply output

Inherit: Inherit via internal rule (default) | Inherit: Keep MSB | Inherit: Match scaling | Inherit: Inherit via back propagation | Inherit: Same as first input | <data type expression>

Choose the data type for the output of the Matrix Multiply blocks inside these subsystems inside the GRU Projected Layer block.

• ForIteratorSubsystem/ResetGateMode/ResetGateMode_after-multiplication/R*h terms

• ForIteratorSubsystem/ResetGateMode/ResetGateMode_before-multiplication/R*h terms

• ForIteratorSubsystem/ResetGateMode/ResetGateMode_recurrent-bias-after-multiplication/R*h terms

The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType. When you select Inherit: Inherit via internal rule, Simulink chooses a data type to balance numerical accuracy, performance, and generated code size, while taking into account the properties of the embedded target hardware.

For a time step t, the Product block computes the product of the recurrent weights (R) and the hidden state at the previous time step (h_t-1). For more information, see Gated Recurrent Unit Layer.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	RecurrentWeightsMatrixMulitplyOutDataTypeStr
Values:	'Inherit: Inherit via internal rule' (default) | 'Inherit: Keep MSB' | 'Inherit: Match scaling' | 'Inherit: Inherit via back propagation' | 'Inherit: Same as first input' | <data type expression>

Input projector weights matrix multiply output — Data type of input projector weights matrix multiply output

Inherit: Inherit via internal rule (default) | Inherit: Keep MSB | Inherit: Match scaling | Inherit: Inherit via back propagation | Inherit: Same as first input | <data type expression>

Choose the data type for the output of the Product block InputWeightsMatrixMultiply/Q_in'*x_full inside the GRU Projected Layer block. The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType. When you select Inherit: Inherit via internal rule, Simulink chooses a data type to balance numerical accuracy, performance, and generated code size, while taking into account the properties of the embedded target hardware.

The Product block computes the product of the transposed input projector weights (Q_in^T) and the input (x_full). For more information, see GRU Projected Layer.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	InputProjectorWeightsMatrixMultiplyOutDataTypeStr
Values:	'Inherit: Inherit via internal rule' (default) | 'Inherit: Keep MSB' | 'Inherit: Match scaling' | 'Inherit: Inherit via back propagation' | 'Inherit: Same as first input' | <data type expression>

Output projector weights matrix multiply output — Data type of output projector weights matrix multiply output

Inherit: Inherit via internal rule (default) | Inherit: Keep MSB | Inherit: Match scaling | Inherit: Inherit via back propagation | Inherit: Same as first input | <data type expression>

Choose the data type for the output of the Matrix Multiply blocks inside these subsystems inside the GRU Projected Layer block.

• ForIteratorSubsystem/ResetGateMode/ResetGateMode_after-multiplication/R*h terms/ProjectHiddenState

• ForIteratorSubsystem/ResetGateMode/ResetGateMode_before-multiplication/R*h terms/ProjectHiddenState

• ForIteratorSubsystem/ResetGateMode/ResetGateMode_before-multiplication/R*h terms/ProjectHadamardProduct

• ForIteratorSubsystem/ResetGateMode/ResetGateMode_recurrent-bias-after-multiplication/R*h terms/ProjectHiddenState

The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType. When you select Inherit: Inherit via internal rule, Simulink chooses a data type to balance numerical accuracy, performance, and generated code size, while taking into account the properties of the embedded target hardware.

For a time step t, the Product block computes the product of the transposed previous hidden state (h_t-1^T) and the output projector weights (Q_out). For more information, see GRU Projected Layer.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter:	OutputProjectorWeightsMatrixMultiplyOutDataTypeStr
Values:	'Inherit: Inherit via internal rule' (default) | 'Inherit: Keep MSB' | 'Inherit: Match scaling' | 'Inherit: Inherit via back propagation' | 'Inherit: Same as first input' | <data type expression>

Execution

Sample time — Option to specify sample time

-1 (default) | scalar

Specify the discrete interval between sample time hits or specify another type of sample time, such as continuous (0) or inherited (-1). For more options, see Types of Sample Time (Simulink).

By default, the block inherits its sample time based on the context of the block within the model.

Programmatic Use

To set the block parameter value programmatically, use the set_param (Simulink) function.

Parameter: SampleTime

Data Types: char

Values: '-1' (default) | scalar

Extended Capabilities

C/C++ Code Generation
Generate C and C++ code using Simulink® Coder™.

Version History

Introduced in R2025a