Emotions are structured biological signals.
TEG-Blue makes them legible.
Open science — transparent research, credited sources, testable claims. Built on 139+ established theories. Open to critique and validation.
When the body can't restore itself, something else steps in.
A single thread runs through all twelve frameworks. Describing a regulation substitute — at a different scale, at a different cost, and a way to reverse the thread — not by adding another substitute, but by building the original.
Empirical evidence
The four-mode gradient has been tested against 10,000+ natural conflict narratives. Here are the key findings.
Empirical Validation — Four-Mode Gradient Framework
Computational analysis of unstructured text. Four regulatory modes reliably detected. De-escalators showed significantly higher rates of self-awareness, perspective-taking, and emotional differentiation. Mode classifications correlated with independent community moral judgments.
What TEG-Blue is
An integrative architecture that maps how nervous system states shape emotional behavior — and what it takes to move between them. Organized as four connected parts:
1 — Measurement
Three Core Models
The Inner Compass (M1) maps where the nervous system is pointing. The Three Awareness Capacities (M2) explain how accurately it reads — and what goes offline when it gets stuck. Biological Restoration (M3) describes the body's designed return path.
Core Models →2 — Explanatory
12 Frameworks
Why these patterns exist, how they escalate from one person into social systems, and what makes repair possible. One mechanism runs through all twelve: when the body's natural return path is missing, something else steps in.
12 Frameworks →3 — Applied
Emotional Tools
Gradient-based scales that map behavioral dimensions from regulated to dysregulated, with observable markers at every point. Empathy, accountability, integrity, boundary dynamics, and more.
Explore Tools →4 — AI Safety
Structured Schemas
A computationally legible translation of the system for AI alignment — designed to replace binary safe/unsafe classifications with contextual, state-aware assessments.
AI Safety →The full architecture, including how the parts connect, is at System Overview. The 139+ source theories are documented at Scientific Foundations.
Open science — no gates, no applications
Everything here is published under CC BY-NC-SA 4.0. No access restrictions. No required collaborations. A researcher can take the framework, the data, and the methodology and work with them independently.