Probabilistic Neural Network Model

This visualization demonstrates a probabilistic linking mechanism to construct neural network architecture. Links between nodes are formed with decreasing probability as layers progress deeper, mimicking synaptic pruning in biological neural networks.

Each layer's connectivity follows its own probability distribution, creating a unique but reproducible network topology that balances density and efficiency. Groups represent functional modules or subsystems, while layers represent processing stages. Multiple functional modules can (and often do) co-exist at the same processing stage, so different groups share layers by design to enable parallel, specialized processing within each stage.

Created by: Alix @ HF Labs

Show Controls ▲
Controls
Mouse:
Zoom: Scroll to zoom in/out
Pan: Hold Alt+drag (or middle button)
Rotate: Click and drag background
Move: Drag individual nodes
Keyboard:
+/- : Zoom in/out
Arrow Keys: Pan view
R/L : Rotate clockwise/counter
H : Reset view • Space: Toggle rotation
F : Focus on hub (exit ring center)
C : Toggle controls panel
Node Categories (Press Key):
1: Entry • 2: Processing • 3: Agents
4: Planning • 5: Evaluation • 0: Exit
6: Feature Extraction • 7: Attention
8: Memory • 9: Decision Matrix