SDCR-CORE

---

Conceptual Overview

SDCR is a symmetry‑selection mechanism, not a modification of quantum mechanics.
Within standard open‑system theory, SDCR:

• identifies symmetry‑aligned operator sectors
• suppresses incompatible dissipative channels by a tunable parameter η
• preserves baseline Lindblad behavior when η = 0
• produces bounded, symmetry‑aligned coherence effects
• enables comparison against fair null selectors

All dynamics remain fully compatible with conventional GKSL formulations.

Learn more:

• SDCR mechanism
• GKSL locked qubit model
• Symmetry selector logic

---

Benchmark Architecture

SDCR‑CORE v0.2 implements the full locked reproducibility pipeline:

1. GKSL Locked Qubit Engine

• Pauli matrices
• ω/2 σ_z Hamiltonian
• baseline + incompatible dissipators
• row‑major vectorized Lindblad superoperators
• deterministic trajectory solver via scipy.linalg.expm

Module: sdcr_core/core/gksl_locked_qubit.py

2. Symmetry Selector

Implements η‑controlled suppression of incompatible channels.

Module: sdcr_core/core/symmetry_selector.py

3. Null Battery (Stage 3)

Includes:

• norm‑matched null
• random‑axis nulls
• channel‑permutation null
• η = 0 recovery baseline

Module: sdcr_core/core/null_battery.py

4. Liouvillian Spectral Audit

Extracts eigenvalues for:

• baseline
• η‑sweep
• target SDCR
• null selectors

Module: sdcr_core/core/liouvillian_spectrum.py

5. Validation Metrics

Reports:

• trace error
• Hermiticity error
• positivity
• η = 0 recovery error
• coherence AUC
• null AUC distribution
• percentile

Module: sdcr_core/core/validation.py

6. Output + Manifest System

Generates:

• CSVs
• JSON summary
• SHA256 manifest
• reproducibility ZIP package

Modules:

• sdcr_core/io/outputs.py
• sdcr_core/io/manifest.py
• sdcr_core/io/package.py

7. Reproducibility Figures

• coherence trajectories
• Liouvillian spectrum vs η
• null AUC histogram
• claim‑state dashboard

Folder: figures/

8. One‑Click Colab Notebook

Public reproducibility path.

Folder: notebooks/

---

Repository Structure

sdcr-core/
  pyproject.toml
  README.md
  sdcr_core/
    core/
      gksl_locked_qubit.py
      symmetry_selector.py
      null_battery.py
      liouvillian_spectrum.py
      validation.py
    io/
      outputs.py
      manifest.py
      package.py
    run.py
  results/
  figures/
  notebooks/

---

Installation

Editable mode:

pip install -e .

---

Running the Full v0.2 Benchmark

One‑click reproducibility pipeline:

python -m sdcr_core.run

This produces:

• results/summary_metrics.json
• results/trajectories.csv
• results/liouvillian_spectrum.csv
• results/null_battery.csv
• results/SHA256_MANIFEST.json
• SDCR_CORE_V02_RESULT_PACKAGE.zip
• all figures in figures/

---

License

MIT License

---

Professional AI Assistance Acknowledgment

This work was completed with assistance from Microsoft Copilot and Google Jules, both of which contributed equally to drafting, structuring, and refining the software and documentation. Their support was used strictly for productivity, clarity, and organization; all technical decisions, implementations, and scientific interpretations remain my own.