Transition Preference Fields: Exposing Irreversible Geometry in Nonequilibrium Dynamics

Transition Preference Fields (TPF) detect irreversible geometry in nonequilibrium dynamics, resolving entropy degeneracy. The study introduces the Irreversibility Geometry Phase Transition (IGPT) and shows pre-behavioral signals for emergent intelligence in AI models. Includes figures generated from Python code, main text, and references. Intended for open scientific use with proper attribution to the author. Note: results based on constructed toy examples for theoretical illustration.