Public BETA v0.5

ATLAS Clarus Interaction Wheel

What happens when one colour is perceived in the context of another?

The ATLAS Clarus Interaction Wheel explores this question using a reproducible computational model. Select a source colour A, a context colour B, and an area relationship. The system calculates a continuous interaction target and searches the complete active ATLAS master of 13,283 references for the closest existing context-appearance reference A′.

New in BETA v0.5 The result is no longer shown on its own. A new diagnostic layer also indicates how robust the selected A′ is.

Target Error How closely does A′ represent the continuous interaction target?

Winner Margin How clearly does A′ outperform the second-best candidate?

Local Density How densely populated is the ATLAS neighbourhood around A′?

Density Stability Does the same A′ remain selected under small controlled target perturbations?

Importantly, these diagnostics do not modify the result. The A′ selection method remains unchanged from v0.4. No density correction is applied.

A remains A.
B remains B.
A′ is a derived context-appearance reference.

The ATLAS Clarus Interaction Wheel is an experimental computational design tool. It is not a physical colour measurement and does not claim to model human colour perception universally.

Identity first. Interaction second. Diagnostics third.