Simulation Report

Generated by octave --no-gui examples/generate_simulation_report.m.

This report uses intentionally small sample counts so it can run quickly in CI or a local terminal. Exact recovery tests live in tests/run_all_tests.m.

Reproducibility Metadata

Field	Value
Generated at	`reproducible`
Octave version	`8.4.0`
RNG seed	`7`
Surface-3 cache version	`surface3-v1`

Code Summary

Code	Physical qubits	Distance / model	Correctness coverage
Bit-flip repetition	3	Distance 3 for X errors	Corrects one X error; Monte Carlo X-noise sweep
Phase-flip repetition	3	Distance 3 for Z errors	Corrects one Z error
5-qubit perfect	5	Distance 3 stabilizer code	Corrects every single-qubit X/Y/Z error
Steane	7	Distance 3 CSS code	Corrects every single-qubit X/Y/Z error
Shor	9	Distance 3 concatenated-style code	Corrects every single-qubit X/Y/Z error
Bacon-Shor	9	Distance 3 subsystem Pauli-frame model	Corrects every single-qubit X/Y/Z error by logical residual parity
Surface-3 prototype	9 data qubits	X/Z/Pauli code-capacity model	Minimum-weight lookup corrects single X, Z, and Y errors
Generic surface layout	odd d-by-d data grids	Code-capacity Pauli model	Exact d=3 lookup, bounded d=5 lookup, bounded graph baseline, and peeling fallback for decoder comparisons

Depolarizing Noise Snapshot

Independent per-qubit depolarizing noise; logical failure is estimated by state fidelity after recovery.

Code	p	Trials	Logical failure
Bit-flip	0.02	40	0.025
Bit-flip	0.08	40	0.100
Bit-flip	0.14	40	0.225
Phase-flip	0.02	40	0.075
Phase-flip	0.08	40	0.200
Phase-flip	0.14	40	0.150
5-qubit	0.02	40	0.000
5-qubit	0.08	40	0.125
5-qubit	0.14	40	0.075
Steane	0.02	40	0.000
Steane	0.08	40	0.125
Steane	0.14	40	0.250

Noisy Syndrome Rounds

3-qubit bit-flip code with data error probability 0.08 and syndrome readout error probability 0.08.

Rounds	Trials	Logical failure
1	400	0.172
3	400	0.058
5	400	0.020

Bacon-Shor Pauli-Frame Snapshot

3x3 Bacon-Shor subsystem-code Pauli-frame model under independent depolarizing noise.

p	Trials	Logical failure
0.02	200	0.010
0.08	200	0.075
0.14	200	0.275

Surface-3 Prototype

Small surface-code model with cached minimum-weight correction lookup for X-only, Z-only, and Pauli noise.

Error probability	Trials	X-only failure	Z-only failure	Pauli failure
0.02	500	0.016	0.006	0.002
0.08	500	0.122	0.098	0.072
0.14	500	0.326	0.202	0.204

Surface-3 channel cache version: surface3-v1.

Surface-3 Noisy Syndrome Rounds

Classical readout noise on syndrome bits with majority vote across repeated rounds.

Rounds	Trials	X-channel failure	Z-channel failure
1	300	0.353	0.387
3	300	0.200	0.137
5	300	0.143	0.080

Surface-3 noisy-syndrome cache version: surface3-v1.

Generic Surface-Layout Distance Benchmark

Variable-distance d-by-d surface-layout model under independent Pauli noise. The table compares the default lookup-plus-peeling decoder against MWPM-style matching and bounded syndrome-graph baselines.

Decoder	Error probability	Trials	d=3 failure	d=5 failure	d=7 failure
min_weight	0.00	8	0.000	0.000	0.000
min_weight	0.03	8	0.000	0.000	0.125
min_weight	0.06	8	0.000	0.000	0.125
min_weight	0.09	8	0.375	0.125	0.250
mwpm	0.00	8	0.000	0.000	0.000
mwpm	0.03	8	0.000	0.000	0.000
mwpm	0.06	8	0.000	0.000	0.125
mwpm	0.09	8	0.375	0.125	0.250
graph_matching	0.00	8	0.000	0.000	0.000
graph_matching	0.03	8	0.000	0.000	0.125
graph_matching	0.06	8	0.000	0.125	0.500
graph_matching	0.09	8	0.375	0.375	0.625

Generated Figures

images/bitflip_logical_error_vs_physical_error.png
images/bitflip_syndrome_distribution.png
images/bitflip_decoder_confusion_matrix.png
images/bitflip_error_weight_distribution.png
images/bitflip_monte_carlo_demo.png
images/qec_recovery_failure_by_error_weight.png
images/qec_depolarizing_logical_error_comparison.png
images/bitflip_noisy_syndrome_rounds.png
images/surface3_logical_error_vs_x_error.png
images/surface3_channel_logical_error_comparison.png
images/surface3_noisy_syndrome_rounds.png
images/surface_distance_logical_error_scaling.png