The Variable Zero Framework A Computational and Thermodynamic Model of Neurobiological Dimorphism

Current neurodevelopmental diagnostic paradigms rely predominantly on observational behavioral models, categorizing neurodivergent phenotypes as pathological deficits. This paper proposes the Variable Zero framework, a theoretical model anchoring these phenotypes in computational neurobiology and the Free Energy Principle (Friston, 2010). We hypothesize a fundamental Neurobiological Dimorphism, characterized by a 67% overproduction of prefrontal neurons alongside temporal neural expansion (Courchesne, 2011) and severely suppressed postnatal synaptic pruning (Tang, 2014) in the neurodivergent (Systemizer) trajectory. Rather than an anomaly, this uncompressed neural architecture is the biological prerequisite for the continuous, bottom-up processing of raw structural data. During the third trimester, the hyper-connected Systemizer fetus utilizes intense epistemic foraging to encode the hydrostatic and thermodynamic absolutes of the womb into a minimum-entropy baseline, designated as Variable Zero. Postnatally, while the neurotypical (Synthesizer) brain executes "lossy data compression" via pruning to prioritize top-down social heuristics, the Systemizer brain retains its expanded hardware and the Variable Zero blueprint. Consequently, the Systemizer predictive engine is computationally compelled to execute continuous background epistemic foraging. Under this framework, the primary clinical phenotypes (e.g., motor stereotypies, tactile seeking, systemizing) are reclassified from behavioral deficits to highly adaptive, closed-loop homeostatic reflexes deployed to synthesize physical boundaries, resolve spatial prediction errors, and manage metabolic heat. By shifting the clinical endpoint from behavioral compliance to thermodynamic equilibrium, Variable Zero provides a mathematically falsifiable model for neurodivergent development.