Intentional Perceptual Attunement (IPA) — Drawing Prototype

A perceptual logic trace + creative trajectory monitor for regulating sense-making in real time

The Intentional Perceptual Attunement (IPA) Drawing Prototype is an interactive research environment for studying how creative perception can be intentionally regulated as a drawing unfolds. Rather than treating drawing as a static artifact or a sequence of strokes, the prototype treats it as a live cognitive process: a shifting field of salience, expectation, and structural tension that changes moment-to-moment. The system provides a canvas for free mark-making alongside real-time instrumentation that visualizes perceptual logic, coherence, and drift—making it possible to observe how attention, structure, and creative direction stabilize (or destabilize) through interaction.

What the prototype is for

In most creative tools, the interface shows you what you drew.

IPA is designed to show you how your perception is moving while you draw.

It is a platform for investigating questions like:

• What does it mean to “notice what feels unfinished” in a measurable way?
  
  

• How does a subtle prompt change what becomes salient next?
  
  

• When does exploration become consolidation—and when does it collapse into drift?
  
  

• How can a co-agent participate without taking over, by responding to the same evolving perceptual field?
  
  

The goal is not to judge drawings, but to make the dynamics of sense-making visible and experimentally tractable.

Core idea: perceptual prompts as control signals

At the center of IPA is a simple but powerful hypothesis:

Perception can be intentionally tuned.

The left panel contains an Active Prompt—for example:
“Notice what feels unfinished.”

This prompt is not meant to produce a sudden change or a forced style shift. Instead, it subtly re-weights salience in the perceptual field. The system treats prompts as attunement operators: gentle control signals that shift what the user is likely to see, prioritize, and resolve next.

In other words, the prompt is part of the cognitive loop, not an instruction layered on top of it.

How it works

The prototype integrates three synchronized components:

1) The drawing canvas (the world)

Users draw freely using mouse or stylus. The environment supports turn-taking: when the user pauses (lifts the pen), the system can give the agent a turn.

2) Perceptual logic trace (the attunement signal)

The Perceptual Logic Trace visualizes the current state of the user’s engagement as a small field monitor. Alongside it are continuously updated measures such as:

• Coherence (how coordinated the unfolding structure is)
  
  

• Drift (how unstable or directionless the process is becoming)
  
  

• Regulation effort (how hard the system/user is working to maintain viability)
  
  

• Participation balance (how influence is distributed across contributors)
  
  

These values are not “scores.” They are intended as process signals—ways of tracking the trajectory of creative regulation as it unfolds.

3) Trajectory monitor (the sense-making curve)

The right panel shows a live trajectory plot—a time-series view of the system’s evolving state. This is a visual representation of your sense-making curve: the rise and fall of coherence, drift, and stability across time as the drawing evolves. It makes phase shifts visible—moments where the process tightens into structure, widens into exploration, or destabilizes into drift.

The co-agent: Kalyri’el as a regulated participant

IPA includes an optional co-agent (“Kalyri’el”) represented through an agent body visualization. The agent does not generate images. It participates by responding to the same trajectory signals and field conditions that the user is shaping.

Its internal state is displayed through interpretable channels such as:

• Coherence
  
  

• Noise
  
  

• Error
  
  

• Drift
  
  

When the agent acts, it “leaves home,” makes a contribution on the canvas, then returns—emphasizing that it is an intermittent participant governed by the same regulation dynamics rather than a constant generator.

This creates a controlled environment for studying human–AI co-regulation: how two contributors coordinate without collapsing the shared process.

Interaction modes: clamp, release, reorg, stall

IPA is designed around the idea that creative activity shifts between regimes. The prototype includes interaction modes that make these regimes explicit:

• Clamp: stabilize and commit to a trajectory (reduce variance, strengthen structure)
  
  

• Release: widen exploration and allow novelty (increase variance, loosen constraints)
  
  

• Reorg: restructure when the current organization fails (reset relations, reframe)
  
  

• Stall: a protective slowdown under rising drift (pause action to prevent collapse)
  
  

These modes allow researchers and users to investigate how regulation operates not just through strokes, but through meta-control over the creative process.

What makes IPA different

Most creative interfaces optimize for output quality or workflow efficiency.

IPA is different in three ways:

1. It treats drawing as a cognitive time-series rather than a static artifact.
   
   

2. It makes perceptual control visible through prompts, traces, and trajectories.
   
   

3. It supports human–AI co-regulation, where the agent’s role is to participate in maintaining viability, not to generate content.
   
   

In short, it is a prototype for studying perception as a regulated process—and for building creative systems that support that regulation rather than replacing it.

Why it matters

If cognition is regulation under drift, then creativity is not just “novel output.”

Creativity is the ability to keep meaning alive while the space of possibilities shifts.

The IPA Drawing Prototype provides an experimental way to observe and shape that process: how coherence forms, how it breaks, and how it can be recovered through intentional attunement.