H-ANS v0.1.0 — Initial Release

Core Architecture

• Clifford algebra engine with complexified velocity map
• FractalMass: 4-fold complex mass with imaginary part = principled uncertainty
• Two-stream design: Schrödinger || Path Integral, fused at CollapseOfLight
• Fractal resonance replaces softmax activation

Build 1: Zero-Day Anomaly Detector

• Pure frequency-domain detection (FFT → Hann → Born collapse → entropy)
• Separation: +0.52, Detection rate: 100%, FPR: 7.5%
• Docker + FastAPI REST API + PCAP capture + Streamlit dashboard

Build 6: BCI Neural Superposition Model

• Clear stimuli → entropy ~0.01 (collapsed)
• Ambiguous stimuli → entropy ~1.38 (superposition)
• Clinical demo with PhysioNet EEG dataset support

Provenance

• Watermarked 6/6 core files with dogtag: 3732a5a3-05c9-43c7-be45-4746df4962b1
• Verify: python verifymy.py

Quick Start

pip install torch numpy scipy
python -m tests.test_sanity
python -m builds.build1_zeroday.detector
python -m builds.build6_bci.cortex

Full Changelog: https://github.com/voltwizeshea/hijoluminic-ans/commits/v0.1.0