# KEY visual grammar

 

Visual language specification for the renderer

This defines how computational state is represented visually without introducing interpretation, bias, or narrative.

 

Parent: KEY presentation intent, KEY renderer spec

Downstream: KEY stage text

 

-----

 

## Purpose

 

The visual grammar defines how different computational states look, how different outcomes are distinguished, how transitions are shown, and how meaning is conveyed without explanation.

 

It does not suggest outcomes, rank candidates visually beyond what the data supports, or introduce visual bias toward success or failure.

 

The visual system must remain epistemically neutral.

 

-----

 

## Governing Constraint (Inherited)

 

The renderer must reveal the computation, not replace it with a story.

 

-----

 

## 1. Core Visual Principles

 

Neutrality: No visual element may imply correctness, incorrectness, importance, or inevitability unless that status is explicitly supported by the computation.

 

Consistency: The same visual rules apply across lens search and composition search. If the underlying kernel is the same, the visual grammar must be the same.

 

Legibility over style: Clarity of state distinction is prioritized over aesthetics.

 

Discrete state transitions: Visual transitions must reflect actual computational transitions. No blending between states. No morphing that implies continuity where none exists.

 

-----

 

## 2. Color System

 

Color encodes state type, not quality or desirability. The color of a candidate says what it is, not how good it is.

 

### Base Palette

 

Active candidate (unscored): Light gray / near-white 

Hard veto (inadmissible): Sharp red / high contrast 

Admissible non-surviving: Desaturated mid-gray, slight decay 

Surviving candidate: Steady mid-tone, slightly brighter neutral 

Promoted structure: Soft blue 

World geometry: Fixed metallic or stone tone 

Accidental / contrast candidate: Same as GIE path 

 

### Critical Color Rules

 

- No color gradient may imply “getting closer to correct” 

- No gold, glow, pulse, or reward-like signaling at any point 

- Success is not visually celebrated — it is shown 

- The promoted structure color (soft blue) must not be applied to candidates during search — it is earned at promotion, not before 

 

-----

 

## 3. Failure Encoding (Critical Section)

 

Two failure types must be visually distinct and instantly readable. This is the most important visual distinction in the demo.

 

### Type 1 — Hard Veto (Inadmissible)

 

Meaning: candidate violates world constraints before invariance scoring. Structurally incompatible.

 

Visual treatment:

- Sharp, immediate removal 

- Brief high-contrast mark at elimination 

- No score shown 

- Contributes only to hard-veto counter 

 

Behavior: appears briefly, is eliminated with no persistence 

 

### Type 2 — Admissible but Non-Surviving

 

Meaning: candidate passed world pre-screen but failed invariance test.

 

Visual treatment:

- Visible in population distribution 

- Score shown below threshold 

- Gradual de-emphasis or decay 

- Remains part of distribution shape 

 

Behavior: persists temporarily, then fades 

 

### Architectural Meaning

 

Hard veto = excluded before evaluation 

Non-survivor = evaluated and failed 

 

-----

 

## 4. Population Representation

 

During search, the population is shown as a distribution, not individuals.

 

Allowed:

- Density cloud 

- Scatter field 

- Histogram of scores 

- Contour projection 

 

Requirements:

- Show spread 

- Show clustering 

- Show convergence 

- Reflect hard-veto as count reduction 

 

Forbidden:

- Highlighting individuals 

- Labeling “best” candidates 

- Showing candidate geometry 

 

-----

 

## 5. Candidate to Structure Transition (Promotion)

 

Meaning: a candidate earns independent standing.

 

Visual behavior:

1. Population clears 

2. Single structure appears 

3. Color shifts to promoted 

4. Structure persists 

 

Requirements:

- Discrete transition 

- No morphing 

- Structure appears as new object 

 

No labels. Only structure.

 

-----

 

## 6. Geometry Representation

 

Raw signal:

- Noisy point clouds 

- No connections 

- Ambiguous 

 

Candidate transformations:

- Show transformed distributions 

- No templates 

 

Promoted structures:

- Clean minimal geometry 

- Stable 

 

Critical constraint:

No recognition cues. No semantic labeling.

 

-----

 

## 7. Composition Stage (High-Risk Section)

 

Core requirement: must feel like the same system on a new substrate.

 

Visual rules:

- Same distribution representation 

- Same failure encoding 

- Same promotion behavior 

- Same secondary panel 

 

Only differences:

- Inputs are structures 

- Geometry type differs 

 

Forbidden:

- Redesigning layout 

- New metaphors 

- Elevated importance 

 

-----

 

## 8. World Geometry

 

Appearance:

- Rigid 

- Fixed 

- Mechanical 

 

Behavior:

- Not visible before evaluation 

- No idle animation 

- Responds discretely 

 

Critical constraint:

The world is indifferent.

 

-----

 

## 9. World Interaction

 

Before:

World is not visible.

 

Failure:

- Constraint A: insertion fails 

- Constraint B: jam 

- Constraint C: no activation 

 

No partial success indicators.

 

Success:

- Clean insertion 

- Rotation 

- Trigger fires 

- State transition 

 

No celebration.

 

-----

 

## 10. Contrast Branch (Critical Constraint)

 

Visual equality is mandatory.

 

The accidental candidate must be:

- equally clean 

- equally plausible 

 

Forbidden:

- roughness 

- dimming 

- distortion 

- predictive cues 

 

Failure must come from the world, not the renderer.

 

-----

 

## 11. Perturbation Test

 

Layout:

Side-by-side, identical framing.

 

Behavior:

- Same world 

- Same perturbation 

- Same visual treatment 

 

Outcome:

Only difference is world response.

 

-----

 

## 12. Motion Rules

 

Allowed:

- Population evolution 

- Promotion appearance 

- World interaction 

 

Forbidden:

- Decorative animation 

- Easing or smoothing  

- Anticipation 

- Idle loops 

 

Motion reflects computation only.

 

-----

 

## 13. Visual Density

 

Only show what is necessary.

 

Avoid:

- clutter 

- redundancy 

- overlapping signals 

 

-----

 

## 14. Hierarchy of Attention

 

1. Active computation 

2. Structural evidence 

3. Stage text 

 

-----

 

## 15. Visual Integrity Tests

 

All must pass:

- Reveals computation 

- No bias 

- No extra information 

- Preserves survival vs appearance 

- Same rules across stages 

- No seams 

 

-----

 

## 16. One-Line Summary

 

The visual system must show what happened, not what it means.

 

-----

 

End of KEY visual grammar


Back to site index