> ## Documentation Index
> Fetch the complete documentation index at: https://docs.cosentriq.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Glossary

> Definitions for common CiCi terms.

**APM** — AI-based participant model used to produce modeled responses.

**Authoritative observation** — The signal observation used for reporting after the applicable mapping or review process completes.

**Behavior Refinement Brief** — A frozen, versioned interpretation and recommendation artifact.

**Evidence basis** — The class and source of evidence supporting a claim.

**Human validation** — Collection, review, and sufficiency assessment of participant responses.

**Implementation Packet** — Copyable intervention instructions, drafts, examples, acceptance criteria, and regression tests.

**MIA** — CiCi's AI analysis and generation layer. MIA-generated hypotheses remain labeled as hypotheses.

**Notebook** — The container for an evaluation program and its product context.

**Scenario** — A shared situation used to compare variants and evidence arms.

**Signal** — A defined behavior or response pattern the evaluation can score.

**Signal Gap** — The human difference between variants minus the APM difference between variants, blocked by shared scenario.

**Sprint** — A frozen execution of an evaluation design.

**Sufficiency** — Whether the collected and reviewed evidence supports the intended claim under the frozen rules.

**Variant** — A product experience or configuration included as one comparison arm.
