Reproducible Qiskit research experiments with saved data, figures, reports, Azure Quantum paths, and explicit claim boundaries.
This repository is the parent suite. Experiment 14, Q-HEAT, is maintained as a standalone publication repo:
https://github.com/Sam-AEC/q-heat-material
| # | Experiment | Qubits | Main artifact | Boundary |
|---|---|---|---|---|
| 01 | Shor toy factorization | 4-8 | toy factoring results | Not cryptographically relevant |
| 02 | Quantum teleportation | 3 | fidelity benchmark | Circuit/protocol benchmark |
| 03 | Grover search | 3-7 | small oracle scaling | No large-scale search claim |
| 04 | VQE molecular energy | 2-4 | exact-checkable energies | Small Hamiltonians only |
| 05 | QAOA engineering optimization | 6-7 | instance-level MaxCut comparison | No broad optimization advantage |
| 06 | Certified randomness | 2+16 | CHSH/randomness workflow | Statistical workflow |
| 07 | Quantum error correction | 1-5 | repetition-code suppression | Noise-model result |
| 08 | Quantum network | 4-5 | swapping/GHZ checks | Idealized circuit workflow |
| 09 | Fermi-Hubbard VQE | 4 | exact-checkable Hubbard scan | Small model only |
| 10 | Hybrid QNN | 4 | MNIST 0-vs-1 benchmark | Local statevector QNN |
| 11 | DQPT Loschmidt/return-rate proxy | 4-8 | finite-size cusp analysis | Proxy observable; emulator path |
| 12 | J1-J2 OTOC butterfly velocity | 4-16 | finite-size OTOC velocity sweep | Statevector/emulator, finite size |
| 13 | Schwinger confinement | 4-8 | finite-size lattice gauge benchmark | Statevector/emulator path |
| 14 | Q-HEAT material screening | 4 | standalone defensive publication | External repo, not duplicated here |
| 15 | Q-MAT facade inverse design | 12 | AI + QUBO/QAOA material-stack search | Local proxy model; no certified material |
quantum-experiments: parent suite and reproducibility archive.q-heat-material: canonical standalone Q-HEAT publication package.quantum-kernel-sensing: separate quantum-kernel benchmark repo.15_qmat_facade_inverse_design: active local Experiment 15 publication package.
See GITHUB_REPO_AUDIT.md for the current repository-overlap audit.
pip install -r requirements.txtRun scripts from the repository root with UTF-8 enabled on Windows:
python -X utf8 11_dqpt_loschmidt/code/run_experiment.py --mode local
python -X utf8 12_j1j2_otoc/code/run_experiment.py --mode local
python -X utf8 13_schwinger_confinement/code/run_experiment.py --mode local
python -X utf8 15_qmat_facade_inverse_design/code/run_experiment.py --mode localWorkspace metadata is stored without secrets:
- Resource group:
AzureQuantum - Workspace:
caribbeanazure - Location:
westeurope
Subscription ID is loaded at runtime from the authenticated Azure CLI session.
Current public scripts target the Quantinuum H2-1 emulator path unless a real .qpu. target is explicitly enabled and recorded in the result artifact.
python -X utf8 submit_azure_batch.py
python -X utf8 retrieve_azure_batch.py --waitReal QPU claims require all of the following in the data artifact:
- target ID containing
.qpu.; - Azure job ID;
- shot count;
- returned counts;
- no fallback to simulator/emulator.
This repo is worked on by multiple AI agents. Read these files before editing:
AGENTS.mdCLAUDE.mdGEMINI.md
For a monitorable Gemini/Claude/Codex council run:
.\tools\run_ai_council.ps1 -PromptFile .\agent_runs\20260520_experiment15_council\council_prompt.mdRaw council logs are ignored. Curated summaries such as 15_AGENT_COUNCIL.md are kept.
Each experiment follows the same structure where possible:
NN_experiment_name/
code/
publication/
report.md
data/
figures/
Experiment 15 current artifacts:
15_qmat_facade_inverse_design/publication/report.md15_qmat_facade_inverse_design/publication/data/qmat_results.json15_qmat_facade_inverse_design/publication/figures/15_qmat_facade_inverse_design/HANDOVER.md
This repository does not claim general quantum advantage.
Simulation, emulator, and real-QPU runs are labeled separately. A material-screening or inverse-design workflow is not a certified material unless supported by physics simulation, prototype fabrication, and lab validation.
@misc{quantum_experiments_2026,
title = {Quantum Experiments: Reproducible Qiskit Research Suite},
author = {Sam-AEC},
year = {2026},
url = {https://github.com/Sam-AEC/quantum-experiments}
}