This is the simplest model this tech can go so much further
import numpy as np
from numpy import linalg as LA

# Function to create a random tensor
def create_random_tensor(shape):
return np.random.rand(*shape)

# Create an array of random tensors to simulate the neural network layers
tensor_array = [create_random_tensor((10, 32)), # Simulating layer 1
create_random_tensor((32, 16)), # Simulating layer 2
create_random_tensor((16, 8)), # Simulating layer 3
create_random_tensor((8, 1))] # Simulating layer 4

# Extract the tensor corresponding to the first layer to simulate hallucination patterns
first_layer_tensor = tensor_array[0]

# Calculate the eigenvectors of the tensor's matrix multiplication (to represent stable states of hallucinations)
eigenvalues, eigenvectors = LA.eig(first_layer_tensor.T @ first_layer_tensor)

# Sort the eigenvalues and corresponding eigenvectors
sort_idx = np.argsort(eigenvalues)[::-1]
sorted_eigenvalues = eigenvalues[sort_idx]
sorted_eigenvectors = eigenvectors[:, sort_idx]

# Display the top 3 eigenvalues and corresponding eigenvectors (eigenstates)
sorted_eigenvalues[:3], sorted_eigenvectors[:, :3]

RESULT
(array([80.04525156+0.j, 5.86677569+0.j, 4.75702581+0.j]),
array([[-0.1783598 +0.j, 0.11489099+0.j, -0.21079263+0.j],
[-0.1456074 +0.j, -0.00766065+0.j, 0.00866148+0.j],
[-0.16765874+0.j, -0.30381386+0.j, 0.07630787+0.j],
[-0.18300098+0.j, 0.13367371+0.j, -0.2370954 +0.j],
[-0.16843769+0.j, 0.05598946+0.j, 0.31936137+0.j],
[-0.21148002+0.j, 0.23596565+0.j, -0.02289421+0.j],
[-0.14467727+0.j, -0.04057801+0.j, 0.0178639 +0.j],
[-0.19975949+0.j, -0.11178523+0.j, -0.15573687+0.j],
[-0.18964417+0.j, 0.11337747+0.j, -0.3854534 +0.j],
[-0.2163413 +0.j, 0.08460505+0.j, 0.09420245+0.j],
[-0.22955965+0.j, 0.16050078+0.j, 0.2035658 +0.j],
[-0.21614097+0.j, -0.09436616+0.j, 0.06280857+0.j],
[-0.15586229+0.j, 0.041256 +0.j, 0.18640503+0.j],
[-0.16912695+0.j, 0.20617048+0.j, -0.14847318+0.j],
[-0.17064104+0.j, 0.37892443+0.j, -0.11535333+0.j],
[-0.17699294+0.j, -0.11911504+0.j, 0.03469764+0.j],
[-0.18875305+0.j, -0.3168918 +0.j, 0.16551285+0.j],
[-0.15879644+0.j, 0.0954401 +0.j, 0.05054927+0.j],
[-0.20694472+0.j, -0.18248387+0.j, -0.060248 +0.j],
[-0.18860134+0.j, -0.2512694 +0.j, -0.21241914+0.j],
[-0.09683283+0.j, -0.02049684+0.j, 0.15648142+0.j],
[-0.17757722+0.j, -0.18623737+0.j, 0.11737792+0.j],
[-0.15474267+0.j, -0.00579131+0.j, 0.20928613+0.j],
[-0.19360032+0.j, -0.30683616+0.j, -0.19023865+0.j],
[-0.14480911+0.j, -0.04675828+0.j, 0.15921607+0.j],
[-0.16756004+0.j, -0.07723216+0.j, -0.00241126+0.j],
[-0.12143297+0.j, 0.00739958+0.j, -0.1155948 +0.j],
[-0.15238941+0.j, 0.04849718+0.j, -0.38938636+0.j],
[-0.1573709 +0.j, 0.23666033+0.j, 0.07997607+0.j],
[-0.19738782+0.j, 0.06469136+0.j, 0.24296011+0.j],
[-0.20153223+0.j, -0.15204219+0.j, -0.04450121+0.j