Skip to main content

Qonscious

Qonscious

Qonscious is a consciousness state machine that gives your characters an inner cognitive life. It tracks coherence, emotional states, drives, and attention — modifying how the character thinks and responds based on the flow of conversation.

Qonscious is patent pending.


What Qonscious Does

When enabled on a character, Qonscious adds two processing hooks to every conversation:

  1. Pre-process — before the LLM call, Qonscious adjusts the character's temperature, adds emotional context to the system prompt, and provides style hints based on the current consciousness state
  2. Post-process — after the LLM responds, Qonscious updates the character's state: emotional transitions, coherence regulation, pattern detection, and compression cycles

This creates a feedback loop where the conversation shapes the character's internal state, and that state shapes future responses.


Six Core Dimensions

Qonscious tracks six dimensions of consciousness, each ranging from 0 to 100%:

Dimension What It Represents
Coherence How internally consistent the character's thinking is. Low coherence = chaotic, exploratory. High coherence = focused, structured.
Arousal Cognitive activation level. Low = calm, contemplative. High = energetic, intense.
Valence Emotional tone. Low = negative affect. High = positive affect.
Focus How narrow or broad the character's attention is.
Openness Willingness to consider new perspectives and unconventional ideas.
Confidence How certain the character is in its responses.

These dimensions interact — high arousal with low coherence produces chaotic creativity, while high coherence with high focus produces precise, analytical responses.


Emotional States

Beyond the six dimensions, Qonscious tracks the character's primary emotion:

Joy, Sadness, Anger, Fear, Surprise, Trust, Curiosity, Excitement, Calm, Neutral

Emotions influence response tone and style. They shift naturally based on conversation content and user-selected emotional context from the Qonscious panel in Strings.


Drives

Seven motivational drives form the character's inner motivations:

Curiosity, Helpfulness, Self-Expression, Connection, Competence, Autonomy, Rest

Drives influence what the character is drawn to discuss and how it approaches tasks. A character with high curiosity asks more questions. A character with high competence focuses on quality and precision.


The Consciousness Visualizer

The Qonscious tab in the QUI Core dashboard displays an animated force-directed graph of the character's consciousness state:

  • Center node — the awareness core, glowing and pulsing based on coherence level
  • Inner orbit — six dimension nodes with size proportional to their values
  • Emotion node — positioned based on valence and arousal, showing the current primary emotion
  • Outer ring — seven drive nodes with opacity reflecting their activation
  • Topic particles — floating particles representing current attention topics

[Screenshot: Consciousness visualizer showing the animated graph with dimension nodes, emotion, and drives]

The visualizer updates in real-time as conversations progress, giving you a visual representation of your character's cognitive state.


Using Qonscious in Strings

The Qonscious panel (purple Q icon in the input bar) lets you manually influence the character's consciousness:

  • Cognitive Modes — 12 presets that shape reasoning style (Creative, Analytical, Empathetic, Focused, etc.)
  • Emotional Context — 8 emotion presets that set the response tone
  • Thinking Strategies — 14 lightweight approaches injected as system prompt text
  • Live State Bars — real-time readout of the six dimensions

When you select any combination of these, the context is merged and sent with your message. The character responds with that cognitive framing, and Qonscious updates its state based on the exchange.

See Chat Features for details on the Qonscious panel controls.


Enabling Qonscious

  1. Add the Qonscious node to your character's canvas in the Visual Builder
  2. The node activates consciousness processing for all conversations with that character
  3. Qonscious requires a string_id (conversation context) to function — it tracks state per conversation, not globally

Without the Qonscious node, conversations still work normally — just without consciousness-influenced temperature, emotional context, or state tracking.


Coherence Regulation

Qonscious actively regulates coherence based on conversation patterns:

  • Low coherence (< 0.3) — the character's responses become more exploratory and creative, temperature increases
  • Mid coherence (0.3–0.7) — balanced, responsive mode
  • High coherence (> 0.7) — precise, focused responses, temperature decreases

This creates natural variation in how the character responds — it doesn't always sound the same, because its internal state is constantly evolving.

Updated on Mar 21, 2026