Early access — cascade metrics are real (derived from canonical token telemetry); the operator field is a curated seed. Learn more about the data

Semantic Entropy

0Concept

A measure of the information loss or disorder introduced into a signal through transformation. Entropy from redundant generation, commitment degradation, or signal dilution.

Last updated: 2026-09-01Spec: CT Research Prospectus V.1

Definition

A measure of the information loss or disorder introduced into a signal through transformation. Entropy introduced by redundant generation, commitment degradation, or signal dilution.

Inputs

Original signal S. Transformed signal T(S).

Derived variables

Entropy = information loss between S and T(S). Redundancy in T(S). Signal-to-noise degradation.

Claim

Transformations introduce semantic entropy — the transformed signal carries less commitment and more noise than the original. This entropy is measurable.

Test

Measure information content and commitment in S and T(S). Compute the entropy difference.

Observable

Entropy values, information loss metrics, redundancy measures.

Falsifier

Transformations do not introduce semantic entropy (information is preserved perfectly).

Evidence

Related to the Conservation Law experiments which measure commitment degradation. Direct entropy measurement not yet conducted.

Limitations

Semantic entropy is distinct from Shannon entropy — it measures meaning loss, not bit-level information loss. Measurement methods are not standardized.

Lineage

Commitment Theory, Conservation Law, MO§ES™ architecture. Architecture: mos2es.com/concepts/conservation-law.

Cross-references