How pref0 learns communication style preferences

Tone, length, format, structure — everyone communicates differently. pref0 learns how each user wants their AI agent to respond and compounds these preferences over time.

What pref0 extracts

Examples of communication style that pref0 detects and scores from conversations.

"Bullet points, not paragraphs"

format: bullet_points0.70
Explicit correction

"Keep it under 5 lines"

length: concise0.40
Explicit instruction

"More casual, drop the corporate speak"

tone: casual0.70
Explicit correction

"Start with the answer, then explain"

structure: answer_first0.40
Explicit instruction

"No emojis"

emojis: never0.70
Explicit instruction

User consistently asks for shorter responses

verbosity: minimal0.30
Behavioral pattern

Confidence progression

See how a single preference compounds in confidence across multiple conversations.

Session 1

"Make it shorter"

0.40
Session 2

"Too long. Bullet points."

0.70
Session 4

"Keep it brief"

0.85
Session 6

Consistently prefers concise responses

1.00

Related preference types

See preference learning in action

Your users are already teaching your agent what they want. pref0 makes sure the lesson sticks.