Ô-HAT Quantum: Detecting Quantum-to-Classical Phase Transitions via a Structural Phase-Space Diagnostic

Overview

This repository contains the code, data, and manuscript for the Ô-HAT Quantum MZI experiment — demonstrating that the Ô-HAT operator robustly detects quantum-to-classical phase transitions in a Mach-Zehnder interferometer across three noise regimes.

Key Results

═══════════════════════════════════════════════════════════════════
                         Coherent ΔH    Decohered ΔH   Δ(ΔH)   d
───────────────────────────────────────────────────────────────────
NumPy ideal             -2.60 ± 0.02   -3.21 ± 0.01   0.61   40.0
AerSim IBM-realistic    -2.69 ± 0.02   -3.21 ± 0.01   0.53   37.0
AerSim artificial T2=1ns -2.60 ± 0.02  -3.19 ± 0.01   0.60   43.0
───────────────────────────────────────────────────────────────────

• Cohen's d > 35 across all noise regimes
• Only 7.5% attenuation from ideal to realistic superconducting qubit noise
• White noise negative control: |ΔH| > 5 in <0.2% of trials

Files

File	Description
paper.md	Full manuscript (6 sections, ~18.5 KB)
qiskit_mzi_ohat.py	Qiskit MZI × Ô-HAT simulation script (NumPy / AerSimulator / Artificial modes)
results/qiskit_mzi_ohat_results.json	Three-layer results (30 trials per regime)

Reproduction

Requirements

pip install qiskit qiskit-aer numpy scipy

Run

# NumPy ideal (no Qiskit needed, ~5 seconds)
python qiskit_mzi_ohat.py --mode numpy --n-trials 30

# AerSimulator IBM-realistic noise (~2 minutes)
python qiskit_mzi_ohat.py --mode simulator --n-aer-trials 30

# Artificial T2=1ns sensitivity analysis (~3 minutes)
python qiskit_mzi_ohat.py --mode artificial --n-aer-trials 30

Citation

@software{ohat_quantum_mzi_2026,
  author       = {DR and MKP},
  title        = {Ô-HAT Quantum: Detecting Quantum-to-Classical Phase Transitions via a Structural Phase-Space Diagnostic},
  year         = 2026,
  publisher    = {Zenodo},
  doi          = {10.5281/zenodo.XXXXXXXXX},
  url          = {https://github.com/MMDR10/ohat-quantum}
}

License

CC BY 4.0 — Free to share and adapt with attribution.

Related Work

• Ô-HAT Caliper — Corporate balance sheet phase-space diagnostic (d=1.80)
• Attention Chunking — LLM context-window structural analysis