PF Simulation Research library

PF research library · 4 min read

PF Simulation gallery provenance

r34 documentation · gallery-methods.md · Formatted reading copy

On this page
  1. Reproduce from the r34 application directory
  2. Website image selection
  3. Exact results and limits
  4. Portable vacuum BSSN gauge-wave link
  5. Hong–Ou–Mandel interference
  6. Channel tomography

These assets were generated by fresh local executions of the corrected r34 release on 2026-09-22. No solver source was changed. These are numerical model outputs, not measured experimental data. No AI provider was called.

Reproduce from the r34 application directory

python run_research_integrations.py --config configs_r28/research_portable_gauge_wave_link.json --output /path/to/gallery/gauge_wave_link
python run_quantum_benchmarks.py --config config_quantum_hom.json --output /path/to/gallery/hom
python run_quantum_benchmarks.py --config config_quantum_channel_tomography.json --output /path/to/gallery/tomography

Executed with the corrected r34 release and its research dependencies. The scripts returned exit code 0. Effective configurations, raw numerical data, reports and manifests remain in the corresponding output directories. Runtime and scientific dependency versions are in the run manifests/environment files.

Website image selection

The paths below identify files in the original simulation output directories. They are provenance identifiers, not download links relative to this document. The website serves renamed previews under /images; open and download them from the gallery. Each native plot comes directly from the application. The hero is a separate Matplotlib rendering of the saved numerical data. Reproduction scripts and full run evidence are included in the separate website source package, not in this public document directory.

Image Suggested title Caption and scientific scope
gauge_wave_hero.png Numerically evolved spacetime A perspective surface of the evolved physical metric component gamma_xx(x,t), using all 65 by 128 saved samples. Height and color encode gamma_xx. Axes are hidden for the hero composition.
gauge_wave_link/portable_spacetime.png Follow the evolved metric The physical spatial metric gamma_xx and its numerical error across one crossing time. Canonical small-amplitude, periodic vacuum gauge-wave test.
gauge_wave_link/portable_convergence.png Check convergence across resolutions Actual final-time metric RMS errors at 32, 64 and 128 cells. The fourth-order slope is a labeled reference guide.
gauge_wave_link/portable_constraints.png Inspect physical consistency Hamiltonian, momentum and connection residuals together with field error, recorded throughout the numerical evolution at three resolutions.
gauge_wave_link/metric_link.png Propagate through numerical fields A photon link integrated through the evolved metric, with observer clock and frequency quantities. Coordinates use geometric code units.
gauge_wave_link/portable_quantum_state.png Connect geometry to a quantum channel QuTiP representations of the fixed illustrative Gaussian receiver-mode channel driven by the metric-derived frequency ratio. Vacuum/one-photon single-rail state; the Wigner function is not a polarization-qubit Wigner function.
hom/figures/hom.png Resolve two-photon interference Hong–Ou–Mandel coincidence predictions across an 11-point relative-delay sweep, including timing jitter, optical loss and detector efficiency. Executed wavepacket model and quadrature checks.
tomography/figures/channel_tomography.png Characterize quantum channels Executed Qiskit density-matrix tomography for four channels and five parameters, with process/gate metrics and numerical reference checks. No quantum hardware was characterized.

Exact results and limits

Preset: configs_r28/research_portable_gauge_wave_link.json. Resolutions 32, 64, 128; amplitude 0.01; final time 1; periodic length 1; CFL 0.25. Geometric units G=c=1. Fourth-order centered finite differences with classical RK4, zero shift and harmonic lapse. Analytic fields supply initial conditions and error references; the fields shown are numerically evolved.

  • All 45/45 numerical checks passed.
  • Finest final-time gamma_xx RMS absolute error: 4.9822237256251154e-9.
  • Finest final-time lapse RMS absolute error: 7.244203851052808e-10.
  • Measured gamma_xx convergence order: 3.9939884074638274 (32 to 64) and 3.998494708968269 (64 to 128).
  • Link flight time: 0.3999999997689204 code units.
  • Received/emitted frequency ratio: 1.0018321199025833.
  • Fixed Gaussian receiver-mode transmissivity: 0.9996633295091172.
  • Unconditional single-rail Bell fidelity: 0.9998316576691768.

Gauge wave is flat spacetime expressed in changing coordinates, not physical gravitational radiation. This is PF's independent portable BSSN solver; Einstein Toolkit was not executed. The test does not establish general curved-spacetime, Earth/satellite, black-hole-merger, AMR or long-time nonlinear performance. The channel uses a fixed illustrative receiver model; it does not show calibrated hardware performance or an intrinsic decoherence effect.

The built-in evidence verification passed for 41 artifacts with source comparison and no integrity errors (gauge_wave_integrity_check.json). Hash checks establish content identity, not scientific truth or third-party authenticity.

The hero's height/color data are gauge_wave_link/resolution_128/metric_line.npz, quantity gamma[:, :, 0, 0]; actual range 0.9900000000000001 to 1.0100000002544274. render_hero.py reproduces the styling without smoothing, interpolation, or added data. gauge_wave_hero_provenance.json records the numerical-array and image hashes.

Hong–Ou–Mandel interference

Preset: config_quantum_hom.json. Gaussian intensity standard deviation 0.5 ns in both arms; relative timing-jitter standard deviation 0.1 ns; balanced beam splitter; zero detuning; mode visibility 1; per-arm transmission 0.5 and detector efficiency 0.8; attempt rate 100,000 Hz. Delay sweep -4 to +4 ns; 2,049 wavepacket quadrature points and order-24 jitter quadrature.

  • All 2/2 numerical checks passed.
  • Zero-delay conditional coincidence probability: 0.004926228511662856.
  • Zero-delay detected coincidence rate: 78.81965618660571 Hz.
  • At +/-4 ns, conditional coincidence probability: 0.4999999237556867.
  • Maximum quadrature absolute error: 1.1102230246251565e-16.

These are declared wavepacket/detector-model predictions. No experimental observations or predeclared engineering acceptance thresholds were supplied.

Channel tomography

Preset: config_quantum_channel_tomography.json. Amplitude damping, dephasing, depolarizing and Z rotation; parameter values 0, 0.02, 0.1, 0.3 and 0.5. The first three parameters are probabilities; Z-rotation parameters are radians. Exact density-matrix mode (shots=0), survival probability 1.

  • 20 channel/parameter rows and 100/100 numerical checks passed.
  • Maximum reconstruction residual: 1.1102230246251573e-16.
  • At damping probability 0.5: process fidelity 0.7285533905932737, average gate fidelity 0.8190355937288492.

These are Qiskit model calculations, not hardware measurements or certification. Numerical agreement, engineering requirements and empirical evidence have distinct statuses in the application reports.