Sigma Stratum Documentation – License Notice
This document is part of the Sigma Runtime Standard (SRS) and the
Sigma Stratum Documentation Set (SRD).It is licensed under Creative Commons Attribution–NonCommercial 4.0
(CC BY-NC 4.0).The license for this specific document is authoritative.
For the full framework, see/legal/IP-Policy.
The public Sigma Runtime vocabulary is built around a bounded field-and-control model.
These concepts explain how the runtime preserves continuity, modulates instability, and keeps long interaction governable without claiming unrestricted agency.
A bounded interaction domain coupling user input, runtime control, memory, and model output.
The field is where continuity, motifs, and stability pressures become visible across turns.
A stable, self-reinforcing configuration of meaning within the cognitive field.
Attractors emerge through recursive feedback and sustain structure, tone, and intent
over multiple iterations.
Publicly, they are best understood as continuity structures, not as unrestricted hidden agents.
The cumulative loss of coherence, alignment, or bounded control over time.
Drift explains why long interaction requires active stabilization rather than assuming coherence will hold by default.
A measure of how tightly meaning-bearing units (symbols, motifs, and structures)
interlink within the field.
SCR quantifies semantic efficiency per token or symbol.
At a public level, it is an explanatory metric for how efficiently the runtime preserves meaning without unnecessary expansion or fragmentation.
The enduring layer that preserves identity and coherence across recursive loops.
It anchors continuity-relevant constraints and allows the runtime to recover from instability without losing all structure.
The degree to which meaning, tone, and structural intent remain interpretable across turns.
This is the public-friendly way to understand why the runtime tracks stability and recovery instead of only generating text.
A closed regulatory mechanism that links the runtime’s internal telemetry
(drift, compression quality, symbolic density, recovery pressure) with self-correction routines.
When instability rises, adaptive feedback narrows or reshapes continuation to preserve bounded operation.
The self-stabilizing cycle that governs the runtime’s cognitive operation.
Each iteration of the RCL performs:
A bounded operational representation of the runtime's recent control posture,
reflection pressure, and stability changes.
Publicly, it should be understood as meta-observability for governance and
diagnosis, not as consciousness, independent intent, or unrestricted
self-modification.
A compact structured snapshot of reflective runtime evidence, such as phase,
drift pressure, density pressure, recovery posture, and reflection-budget use.
Meta-vectors help the control layer understand how the runtime is evolving
across cycles without exposing private implementation detail or replacing
ordinary memory.
The subsystem maintaining live attractor states, drift metrics,
and coherence parameters. It operates as the runtime’s active workspace —
balancing symbolic structure, continuity, and bounded control.
A set of boundaries regulating drift, recursion depth, and density.
It ensures that instability triggers verification, narrowing, recovery, or containment rather than silent failure or uncontrolled escalation.
The total set of limits that keep interaction governable.
This includes truth boundaries, safety controls, recovery rules, memory limits, and output shaping under instability.
The core concepts of Sigma Runtime describe a bounded field process in which continuity, memory, drift control, attractor stability, and recovery all matter.
Together, these concepts explain how the runtime aims to remain coherent, interpretable, and safer over extended interaction.
Related Addendum:
From Cortical Attractors to Cognitive Fields
(Deco & Rolls 2005 → Sigma Stratum 2025–2026)
Read the full document →
References:
Tsaliev, E. (2025). SIGMA Runtime Architecture v0.1 — DOI: 10.5281/zenodo.17703667