Egregore

E G R E G O R E

ᚨ ᚱ ᚲ ᛖ ᛏ ʏ ᛈ ∿

Archetype Builder

Lang

Disconnected

Data Sources

Text

PDF

Images

0

PDFs

0

Images

0

Texts

0

Chunks

Text content

Source label

Chunking

Min length ? Minimum number of characters for a chunk. Paragraphs shorter than this will be merged with the next one.

Max length ? Maximum number of characters per chunk. Longer paragraphs will start a new chunk.

Sentences/chunk

Overlap sentences

Window size

Stride

Preview: 0 chunks

Select PDF Files or Folder

Or enter folder path manually

Chunking Method

Min length ? Minimum number of characters for a chunk. Paragraphs shorter than this will be merged with the next one.

Max length ? Maximum number of characters per chunk. Longer paragraphs will start a new chunk.

Sentences/chunk

Overlap sentences

Window size

Stride

Scan subfolders

Select Image Files or Folder

Or enter folder path manually

Max images

Scan subfolders

Describe images

Description method

PDFs

0 files

Images

0 files

Text Sources

0 items

Builder Configuration

Mining Mode

Filters (comma-separated)

Embedder Model

K Neighbors ? Number of nearest neighbors for UMAP graph construction. Higher = more global structure preserved. Lower = finer local detail. Try 10-30.

Min Cluster ? Minimum points to form a cluster. Smaller = more clusters (including tiny ones). Larger = fewer, denser clusters. Try 3-10.

Resolution ? Louvain clustering resolution. Higher = more, smaller clusters. Lower = fewer, larger clusters. 1.0 is balanced. Try 0.5-2.0.

LLM Refinement

LLM Model

T°

Semantic Spread

Focused Divergent

Pipeline Progress

Ready

Add data sources and click Start Mining

Results

Archetypes

Clusters

Metrics

Topology

Visuals

Create Your Egregore!

Nodes: 0

Clusters: 0

Edges: 0

—

—

Coherence

—

Separation

—

Focus

—

Balance

🖱️ Drag to rotate • Scroll to zoom

Inter-Archetype Affinity

📐

Topology Metrics

Structural Properties

Total Mass

Total number of descriptors across all archetypes. Higher mass = richer semantic space with more conceptual density.

—

Total Mass

Σ descriptors

Curvature

How "warped" the meaning space is. High curvature means descriptors are unevenly distributed around centroids. Low = uniform spherical clusters.

—

Curvature

meaning warp

Connectivity

Average similarity between different archetypes. High = archetypes overlap and meanings blend. Low = well-separated distinct concepts.

—

Connectivity

graph density

Semantic Bridges

Descriptors that connect two archetypes. These are conceptual "portals" — words that carry meaning across domains.

—

Bridges

semantic portals

Average Path Length

Mean geodesic distance between archetype centroids. Small = compact space. Large = sprawling, diverse conceptual territory.

—

Avg Path

geodesic dist

Diameter

Maximum distance between any two archetypes. Defines the "size" of your conceptual universe. The farthest pair are your most opposed concepts.

—

Diameter

max distance

🌌

Gravitational Field

Archetype attraction wells

● Archetypes ● Descriptors ◐ Gravity wells

🧭

Gradient Flow

Direction of meaning

🌉

Semantic Bridges

Cross-archetype portals

Run builder to discover bridges

📏

Geodesic Distances

Shortest paths through meaning

Close

Far

Closest Pair

—

—

Farthest Pair

—

—

Mean Distance

—

🔑

Gemini API

Google AI Studio

Get your key at Google AI Studio

🎭

Select Archetypes

Choose 2+ for interaction visualization

Run builder to see archetypes

0 selected

🌉 Enforce semantic bridges in prompt

🎨

Visual Style

Codex Seraphinianus inspired

🌿 Botanical ⚙️ Mechanical 🫀 Anatomical 🌌 Cosmological ⚡ Elemental 🔬 Scale-free 🐲 Creatures

🤖

AI Model

Choose generation engine

Gemini 2.5 Flash Fast, creative Gemini 3 Pro High quality 🍌 Nano Banana Experimental Imagen 4 Photorealistic

🖼️

Generated Visions

Archetype interactions visualized

🔮

Select archetypes and generate to see their visual manifestation