🧠 ABSOLUTE SYNTHETIC INTUITION (ASI)

Modular Prompt System for Z Image Turbo

What is ASI?

ABSOLUTE SYNTHETIC INTUITION (ASI) is a modular prompt architecture designed to structure image generation through logic, weighting, and separation, instead of relying on monolithic prompts.

ASI is a system of wildcards, not a single prompt.

It is optimized for Z Image Turbo and focuses on:

• stability

• efficiency

• reproducibility

• clean subject dominance

🚀 Quick Start (Important)

If you only want to use ASI, you only need one file:

✅ Master Wildcard (Entry Point)

__ASI-Master-Orchestrator__

or the Turbo-optimized version:

__ASI-Master-Orchestrator-v2__

This master wildcard automatically assembles:

• style handling

• logic modules

• render descriptors

• weight logic

• optional negatives

All other files are internal modules used by the orchestrator.

🎯 ASI Design Principles

ASI is built on three core principles:

1. Modularity

Each function is isolated into its own wildcard:

• style

• logic

• render

• isolation

• depth

No monolithic prompt blocks.

2. Dichotomy Weight Logic

ASI uses a strict separation between:

• high-impact weights

• low-impact weights

There are no ambiguous mid-values.
This ensures predictable behavior across generations.

3. Model Awareness

ASI is explicitly designed around the behavior of Z Image Turbo, including:

• token sensitivity

• style bleed tendencies

• depth compression

🧩 System Structure (Overview)

Shared Core

• __ASI-Weight-High__

• __ASI-Weight-Low__

• Style libraries (Front / Back)

These elements are reused across all ASI pipelines.

Generation 1 – Classic Modular

• Fully modular

• Multiple render stacks

• Designed for flexibility and experimentation

Generation 2 – Turbo Optimized (Recommended)

• Reduced token count

• High-impact clustering

• Background style weighted at 0.7

• Strong subject dominance and 2.5D separation

🔍 Core ASI Logic Modules

ASI Logic – Isolation

Prevents:

• color bleeding

• style contamination

• flat subject/background merging

Ensures clean layer separation.

ASI Logic – PopOut

Creates:

• perceived depth

• subtle relief

• foreground dominance

• cinematic subject presence

Especially effective for character-focused images.

🎨 Style Handling (ASI Concept)

• Style-Front defines the primary visual language

• Style-Back reinforces coherence at reduced strength

This avoids overpowering the subject while maintaining consistency.

❌ Why There Is No Mandatory Negative Wildcard

ASI intentionally avoids large negative prompt blocks.

Reasons:

• Z Image Turbo responds better to positive structural guidance

• Most common failure modes are prevented by:

  • isolation logic

  • weight separation

  • controlled style reinforcement

Minimal negatives can still be added manually if desired.

📂 Index File (Documentation Only)

The ASI index file is:

• a human-readable overview

• not part of the generation pipeline

• included purely for orientation and reconstruction

It has no functional impact.

🧪 Intended Use Cases

• Character design

• Concept art

• Cinematic portraits

• Modular style experimentation

• Fast iteration workflows

🏁 Final Notes

ASI is designed to be:

• extended

• modified

• adapted

It is a framework, not a closed system.

If you build on it, that is expected behavior.

💬 Feedback & Iteration

If you create:

• custom modules

• alternative pipelines

• optimizations for other models

feel free to share them.

ASI evolves through use.

Pro-Tip: Workflow Optimization & Roadmap

Pro-Tip for Maximum Efficiency: Let the AI handle the structural heavy lifting. When using an LLM to write your prompts, simply provide the full list of available wildcards. The AI will automatically distribute and position them to achieve the best possible balance based on your specific focus. This automated orchestration ensures the most stable and high-quality results within the ASI framework.

Background & Roadmap: This system is the direct evolution of The Loom - Quality, Style & Perspective and is currently my top recommendation for prompt architecture.

• Status: Currently in Beta (fully functional and highly stable).

• Release: The final integration is scheduled for late January as a major update to the main module: The Loom – Modular Prompt Engine (All-in-One Edition).

• Future Updates: The final version will feature significant refinements, specifically focusing on advanced glow effects and enhanced light-logic.

Pro-Tip: Style Handling & LoRA Compatibility

Multi-Style Synergies: The wildcards are designed to combine multiple visual styles seamlessly. This architecture works best in its standard version, where the internal logic maintains perfect balance.

LoRA Interaction: Note that using external LoRAs will generally not cause significant style shifts or drastic changes to the aesthetic. The ASI framework is built for high structural integrity, making it resistant to style bleeding.

• Exception: High-impact "Punk" styles (e.g., Cyberpunk, Steampunk) are the only categories where LoRAs will exert a dominant influence on the final output.