A translational cross-modal control-cost framework for executive breakdown

Figures

• 

  Figure 1. Hierarchical control architecture and selection power budget. (A) Three-part control scaffold: candidate policy space, active state, and slow constraints, linked by a resource-bounded selection/gating mechanism. Dashed feedback arrow indicates how slow constraints shape future selection. (B) Selection-budget diagram illustrating how the “knowing-doing gap” emerges when the red bar (β · Ψf) plus orange bar (γ · Snoise) exceeds the green bar (α · Psel), progressing from healthy function through stressed operation to clinical breakdown. Final artwork should keep these exact symbol labels for one-to-one correspondence with Equation 5.

• 

  Figure 2. Cross-modal operationalization map for executive friction (Ψf). Central latent construct with four proxy classes. Core validation chain: behavioral, linguistic, and low-cost physiology (HRV/SCR/cortisol). Expansion chain: ROS biomarkers and neural proxies (PCI, Fisher information, theta-gamma coupling, basal ganglia gating). Arrows indicate predicted direction of association; “(inv)” denotes inverse coding.

• 

  Figure 3. Proposed oxidative feedback loop linking sustained executive friction to clinical features. Schematic diagram showing a proposed loop from sustained high Ψf through elevated metabolic demand, increased ROS production, oxidative damage, inflammatory signaling, reduced selection power (Psel), and a deepened knowing-doing gap (dashed red feedback arrow). Dashed boxes at bottom indicate three possible modulation points: pharmacological (↓ Ψf), behavioral (↑ Psel), and antioxidant (↑ clearance).

• 

  Figure 4. Schematic predicted Ψf proxy signatures across clinical conditions. Radar chart showing schematic relative profiles for healthy controls (green), major depressive disorder (red), obsessive-compulsive disorder (purple), and Parkinson's disease (orange) across ten proxy measures. Higher values indicate greater impairment or higher executive friction. Values are schematic relative predictions, not empirical estimates.

• 

  Figure 5. Core-chain protocol and critical-load prediction. (A) Low-cost protocol for H72 and H-NEURO-EXEC-01, integrating two prompt contexts (neutral factual retell vs self-decision narrative), behavioral battery, HRV/SCR recording, appraisal ratings, and NLP extraction for latent-factor estimation. Baseline load is estimated as ρ*=z(Ψfbaseline, latent)-z(Pselproxy) (Equation 7c) in primary analyses; ratio-based ρ is sensitivity-only. (B) Predicted task-switch cost as a function of normalized load ρ: the saturation model (solid red) includes a critical point ρc and nonlinear escalation into a “know but can't do” zone, while the additive model (dashed gray) predicts linear increase.