Skip to content

iDense — Gnostic Estimating Weight Dense Layer

Dense layer that applies gnostic estimating weights (gnostic_weights_i) before the affine transformation.


Overview

iDense extends Dense by multiplying the input by gnostic estimating weights computed from the current batch. This emphasizes the estimation-side weighting path (fi-based weights) before the standard y = x @ W + b affine transform. It inherits all Dense parameters and adds S (scale) and lel (local estimate of location) for the gnostic computation.


Key Features

  • Applies gnostic_weights_i() before dense transform
  • Configurable scale S and lel
  • Inherits full Dense parameter and method interface
  • Supports training mode flag

Parameters

Parameter Type Default Description
in_features int Required Input dimension (n_in)
out_features int Required Output dimension (n_out)
S float | str 2.0 Scale parameter for gnostic weight calculation
lel float None Local Estimate of Location; computed as median if None
weight_init callable XavierUniform Weight initializer
bias_init callable Zeros() Bias initializer

Methods

forward(x, training=True)

Compute estimating weights from x, reshape for broadcasting, multiply into input, then apply dense affine transform.


Example Usage

from machinegnostics.magnet import iDense
layer = iDense(in_features=2, out_features=1, S=2.0)
print(layer(np.array([[1.0, 2.0]])).shape)  # (1, 1)

Author: Nirmal Parmar