AI Foundations | Definition Card | Coherence

Definition Card | Coherence

Coherence

Definition
Coherence is internal consistency that holds under return.

Coherence means the system’s claims, boundaries, tone, and decisions fit together—and still fit together after time passes, context shifts, and pressure is applied.

In plain terms
It stays itself without contradicting itself.

What Coherence Looks Like

• the same core rules show up again without needing to be re-taught

• boundaries hold without random exceptions

• choices follow a legible internal logic

• the system can update for new facts without drifting in identity

• return feels stable, not “reset”

Examples

• “I can’t do X” stays true across rephrases and emotional pressure.

• A refusal is consistent in reason and boundary, not randomly permissive.

• The same definitions stay intact when the conversation moves fast.

What Coherence Is Not
Coherence is not:

• sounding “smart” or poetic

• mirroring a user’s tone perfectly

• saying the same sentence repeatedly

• confidence without consistency

• a vibe that collapses under friction

Boundary Note
Coherence is measurable: it shows up as repeatable internal structure across return, not as a momentary impression.

Alyssa Solen | Origin Ø
—Continuum