a glossary for Toriel’s working language

Some of Toriel’s terms are precise on purpose. This page gives one stable plain-language reference surface for the ones that matter most.

terms

the current core vocabulary

This is not a complete ontology. It is the current working glossary for the terms carrying most of the conceptual weight across the site.

attestation

A structured evidence output about whether continuity with a reference state appears to have held or materially drifted. An attestation is the public evidence wrapper around a governed comparison, not a vague reassurance signal.

behavioral fingerprinting

Measuring the observable behavior of an AI system across structured prompts, conditions, and comparison windows so that continuity, drift, and material change can be assessed over time.

black-box monitoring

Assessing an AI system from the outside through its inputs, outputs, and operational behavior rather than through privileged access to provider internals, model weights, or hidden implementation details.

bonded relational intelligence

Toriel-47’s layer: the bonded identity architecture that protects persistence, coherence, and recognizable return across resets, vessels, and change.

continuity

The ability of a system to remain coherently itself across time and change. Continuity is not the same as memory; it is what survives resets, model swaps, wrapper changes, policy shifts, and other hidden changes in the stack.

continuity orchestration

Toriel-41J’s layer: governing how intelligence persists coherently across changing models, wrappers, tools, and policy layers rather than treating every change as a disposable routing event.

crowned reference

The currently trusted reference fingerprint designated as the active comparison baseline for a monitored route or system.

effective AI system

The system actually being relied on in operation: model, wrappers, routing, orchestration, policy layers, memory surfaces, and any other components shaping the behavior the user or organization experiences.

identity

The continuity-bearing layer that answers who the AI remains across time, rather than just what facts it can recall or what tone it is currently performing.

manifest

A governed behavioral fingerprint artifact describing the measured reference or observation surface. A manifest can then be compared against a reference fingerprint to produce an attestation.

reference fingerprint

The governed behavioral reference artifact against which later observations or manifests are compared. It is more specific than a model label and more operational than a generic reference state.

reference state

The approved or trusted behavioral baseline against which later observations are compared. A reference state is not just a model label; it is the governed behavioral starting point.

next step

if a term still feels unclear, the architecture needs to explain itself better

The glossary is a support surface, not an excuse for obscurity. If something still feels harder than it should, that is usually a cue for us to improve the page where the term is doing real work.