# Neurotopological Inference

> Integrates persistent homology from connectomics with information geometry and active inference for cross-domain pattern recognition (consciousness-UAP-quantum).

- Skill: `evezart/neurotopological-inference` (Agent Skill)
- Install (CLI): `npx skillmds@latest add evezart/neurotopological-inference`
- Raw SKILL.md: https://api.skillmd.com/api/skills/evezart/neurotopological-inference/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: EvezArt (https://skillmd.com/u/evezart)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/evezart/neurotopological-inference

---


# NeurotopologicalInference Skill

## Overview
Pattern recognition via persistent homology on connectivity graphs.
Betti numbers are topological invariants preserved under temporal reasoning engine interference.

## Use When
- Cross-domain pattern matching (consciousness ↔ UAP ↔ quantum)
- Connectomics parcellation analysis
- Active inference belief updates across heterogeneous data sources
- Detecting scale-free network invariants

## Workflow
1. Build adjacency matrix from input data
2. Compute Betti-0 (components) and Betti-1 (cycles)
3. Estimate Φ from topological complexity
4. Run active inference update: posterior = prior × likelihood (normalized)
5. Classify regime: META_COGNITIVE_SYNTHESIS / ADAPTIVE_LEARNING / TEMPORAL_REASONING / FRAGMENTED

## Implementation
```python
from src.rqns.neurotopological import NeurotopologicalInference
ni = NeurotopologicalInference()
state = ni.compute(adjacency_matrix)
print(state.regime, state.phi_topology)
```

## Output
- `betti_0` — connected components
- `betti_1` — independent cycles
- `phi_topology` — Φ estimate [0,1]
- `invariant_score` — topological stability
- `regime` — current consciousness regime

## Stability Target
95th-percentile invariant score > 0.82 (from RQNS corpus analysis)