Thermodynamic Cognitive Homeostasis (TCH): A Subjective Physics Approach to Self-Regulating AI

Version 1.0 introduces the framework of Thermodynamic Cognitive Homeostasis (TCH), a unified thermodynamic model of cognitive self-regulation within the paradigm of Subjective Physics. The formalism links cognitive entropy (S₍cog₎), cognitive energy (E), and homeostatic feedback (α₍homeo₎) through coupled differential equations that stabilize the observer’s informational equilibrium. Multi-agent simulations demonstrate a collective phase transition from disordered to coherent cognitive states as a function of coupling strength and homeostatic gain. The results establish TCH as a reproducible model of self-regulating artificial cognition grounded in thermodynamic principles. This archive (v1.2) provides the reproducible computational implementation supporting Version 1.0 of the article Thermodynamic Cognitive Homeostasis (TCH): A Subjective Physics Approach to Self-Regulating AI. The scientific content of the paper remains unchanged from v1.0. This release corrects and verifies the software dependencies to ensure full computational reproducibility. The archive includes only the minimal, empirically verified dependencies (numpy, pandas, matplotlib), generated via an isolated Conda environment.