Neuromimetic Semantic Model
A compact System 1 sequence model with resonance, activity regulation, and optional fast-weight plasticity.
Overview
The NeuromimeticSemanticModel packages ARCANE's core adaptive mechanisms into a smaller architecture for semantic sequence tasks. Use it when you want a shorter training path, lower default depth, and direct access to resonance and activity regulation.
Usage
neuromimetic-model.py
from gpbacay_arcane import NeuromimeticSemanticModelmodel = NeuromimeticSemanticModel( vocab_size=1000, enable_inference_plasticity=True, use_hierarchical_homeostasis=True, homeostasis_config={"homeostasis_mode": "hybrid"},)model.build_model()model.compile_model()# Standard trainingmodel.model.fit(X_train, y_train)# Current-input refinement with fixed learned weightspredictions = model.predict_with_resonance(X_test, resonance_cycles=2)Comparison vs Foundation Model
| Feature | NeuromimeticSemanticModel | FoundationModel |
|---|---|---|
| Resonance Levels | Compact hierarchy | Configurable deeper hierarchy |
| Training Speed | Lower default depth | More recurrent levels |
| Best For | Compact semantic workloads | Deep hierarchy and adaptation workflows |