Skip to content

Efeatagul/WNEURA

Repository files navigation

🧠 WNEURA v1.2: Multi-Dimensional Cognitive Simulation Platform

WSharp Python Status Focus

Status: Phase 2 (Bridge System & Neurochemistry Ready) 🌉
Field: Computational Neuroscience / Neurochemistry / Bio-Inspired AI
Integration: Optimized for WSharp (C#) & External Control

WNEURA is a neuroscience-focused platform designed to simulate biological brain development, neurochemical balances, and decision-making processes within a multi-dimensional digital environment.

Version 1.2 introduces a "Neurochemical Modulation" (Dopamine/Serotonin) layer, a "Hippocampal Memory System", and transitions to a "Headless" engine architecture, enabling full integration with external software (specifically WSharp) via JSON protocols.


🛠️ Core Engine Architecture

The platform operates through the mathematical integration of three primary layers: Structural, Chemical, and Memory.

1. Structural Layer (The Brain)

Mechanism Analog Function
Amygdala Stress Engine Generates Cortisol based on prediction errors (Surprise).
Agency ($W$) Volition Weight Manages the belief in causality between action and outcome.
Striatum Learning Gate Stops learning (Freezing) if Agency is too low, even if a reward is present.
Hippocampus Memory Center New: Stores memories based on emotional weight (Synaptic Tagging).

2. Chemical Layer (Neurochemistry) - New in v1.2

Neurotransmitter Domain Simulation Mechanism
Dopamine (DA) Motivation & Reward Simulates receptor sensitivity (Tolerance) and addiction cycles.
Serotonin (5-HT) Mood Stability Suppressed by stress; replenished by successful action.
Noradrenaline (NE) Focus (Arousal) Affects performance based on the Yerkes-Dodson law.

📂 Project Structure

The project is organized into core engine components and experimental scenarios:

WNEURA/
├── wneuraa/                 # Core Package (Renamed)
│   ├── __init__.py          # Package initializer
│   ├── agent.py             # Neurological Agent (Decision Maker)
│   ├── brain.py             # Biological Engine (Cortisol/Agency Dynamics)
│   ├── hippocampus.py       # Memory System (Synaptic Tagging) 🧠
│   ├── neuromodulator.py    # Chemistry Lab (Dopamine/Receptor) 🧪
│   ├── config.py            # Hyperparameters
│   └── runner.py            # CLI & C# Bridge Commander
│
├── experiments/             # Scenarios & Tests
│   ├── experiment_lab.py    # The Grand Experiment (Burnout Sim) 📉
│   └── therapy.py           # Rehabilitation Tests
│
└── README.md                # Documentation

Integration & Usage To trigger the WNEURA engine externally (via Terminal or WSharp), use runner.py.

  1. Standard Simulation (CLI)
py wneuraa/runner.py --steps 100 --scenario chaos --output result.json
  1. The Grand Experiment (Burnout Lab) 🔥 To observe Burnout Syndrome, Dopamine tolerance, and the Recovery process live:
py experiments/experiment_lab.py

This command launches a live biological dashboard with ASCII graphics in the terminal.

Validation Experiments The biological accuracy of the model has been proven through four fundamental experiments:

  1. Hysteresis Proof (Permanent Damage) It has been proven that even if stress (Cortisol) is removed, the Agency level does not recover spontaneously.
Observation: Post-trauma, the system remains locked in "Helplessness" mode.
  1. Contingency Switch (Opportunity Blindness) Even when a massive reward (+10) is introduced to the environment, the helpless agent fails to notice this opportunity and cannot update its motivation (Outcome Insensitivity).

  2. Therapy & Rehabilitation Tested via experiments/therapy.py.

Observation: Giving standard rewards to an agent with zeroed Agency (Agency=0) is ineffective. The system only begins to respond when the "Repair Rate" is increased via external support (simulating medication or intense therapy).
  1. Receptor Downregulation (Tolerance & Addiction) - New Tested via experiments/experiment_lab.py.
Observation: During "The Hustle" phase (High Stress + High Reward), although dopamine levels skyrocket, Receptor Sensitivity decreases. Consequently, the agent cannot derive satisfaction even from rewards, eventually leading to Burnout.

Experimental Results Experiment File / Visual Status

Hysteresis      results/figure_01_hysteresis.png         Successful ✅
Dissociation    results/figure_02_dissociation.png       Successful ✅
Contingency     results/figure_03_contingency.png        Successful ✅
Therapy Sim     experiments/therapy.py (Log)             Successful ✅
Burnout Lab     experiments/experiment_lab.py (Live)     Successful ✅

Future Roadmap

[x] Phase 1: Core Engine & Validation (Completed)
[x] Phase 2: Headless Architecture & JSON Bridge (Completed)
[x] Phase 2.5: Neurochemistry & Receptor Dynamics (Completed) 🧪
[x] Phase 2.8: Hippocampal Memory & Tagging (Completed) 🧠
[ ] Phase 3: Full WSharp (C#) Integration and UI.
[ ] Phase 4: Sleep & Replay (Memory Consolidation).
[ ] Phase 5: Multi-Agent Interaction (Social Helplessness).

Developer: [Efeatagul] License: MIT License. Open for scientific use and development.

About

A neurology-focused simulation platform aiming to computationally simulate the brain's development and decision-making processes in a multi-dimensional and multi-factorial manner

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages