Project Title: Source Code for Adversarial Active Inference and Cognitive Destabilization

Summary: This repository contains the computational proof-of-concept for the Adversarial Active Inference (AAI) framework. The provided Python script implements a 3-level Hierarchical Gaussian Filter (HGF) to simulate the endogenous modulation of precision parameters. Key Contents: HGF Simulation (aai_hgf_simulation.py): A standalone Python script that reproduces the Variational Free Energy (VFE) trajectories shown in Figure 2 of the manuscript. Regulatory Logic: The code explicitly models Algorithm 1, contrasting unregulated "Zen Sickness" (Scenario 1) with threshold-gated "Cognitive Annealing" (Scenario 2). Metastable Regime: Demonstrates how the system maintains a bounded, high-energy state (120-150 VFE range) to facilitate second-order structural learning. Instructions: Requirements: Python 3.x, NumPy, and Matplotlib. Run python aai_hgf_simulation.py to generate the simulation plot.