Systemic Inertia in Human Cognition: Drivers of Continuous Representational Growth

This repository contains the Colab-ready implementation of the "Self-Expanding Mind" V4 model, a computational framework integrating Systemic Inertia (SI), Curiosity, and Dream-based Generative Replay. The V4 model formalizes a structural Meta-Prior governing continuous cognitive and representational expansion within an Active Inference architecture, providing a unified simulation of how curiosity-driven exploration during wakefulness and offline reorganization during sleep jointly contribute to long-term structural growth. The repository includes a 1000-epoch simulation exploring continuous representational growth (C_sim), verifying core claims of the V4 model: the structural necessity of Generative Replay for C_sim expansion, optimal tuning of SI, and non-linear interactions among latent variables. Ablation and regression analyses are included to demonstrate cyclical interdependence and confirm theoretical predictions. Included Files: Supplementary_Code.ipynb : Colab notebook containing full V4 system code, 1000-epoch simulation, and enhanced visualizations. Supplementary_Data.csv : Output CSV of the main experiment, including C_sim, H, D, C_dream, and SI. Supplementary_Requirements.txt : Python package dependencies for Colab execution. Supplementary_ReadMe.txt : Instructions and description. Instructions: Open Supplementary_Code.ipynb in Google Colab. Enable GPU runtime for faster computation (optional but recommended). Run all cells sequentially; visualizations will be produced and simulation_results.csv saved. The notebook is fully self-contained; no external data is required.