A discriminating K-ablation against published empirical Clinton–Gore joint distributions, executed end-to-end on ibm_fez (IBM Heron r2). Architectural discrimination with hardware-archived statistical significance.
Plain English
Start here for the narrative, the K-ablation table, and the bootstrap discrimination result. Follow the chain below if you audit outcomes via job IDs.
QQ v3 evidence bundle in the QQ workspace folder. Include copies with a reviewer submission bundle when downloads are required.Navigation: Home · Highlights
The task asks whether QPC's polycontextural architecture (multiple coexisting contextures with their own quantum-logical states) reproduces the empirical structure of contextual cognitive data better than a faithful non-polycontextural control of equal quantum resources. The empirical target is the Clinton–Gore question-order experiment of Wang and Busemeyer (2013). A 1997 Gallup poll of 1,002 respondents found a robust, replicable order effect: the joint answer distribution differs depending on whether the Clinton question is asked first or second. Wang and Busemeyer prove that this empirical structure sits outside any single Kolmogorov probability space (the data satisfies the QQ equality that Bayesian and Markov models miss). That classical theorem is the published backdrop. This pilot measures architectural discrimination on IBM hardware against those joints.
What this pilot delivers:
Empirical target: Wang and Busemeyer 2013, Topics in Cognitive Science 5(4):689–710, Table 1, "Consistency" column. The poll was conducted Sept 6–7, 1997; half of 1,002 respondents answered the Clinton honesty question first, the other half answered the Gore question first.
Contexture-to-qubit mapping is proprietary. Public evidence: K-ablation metrics and IBM job IDs.
v3 was tested for K-ablation discrimination on a noiseless simulator before hardware submission. The simulator pre-flight produced TV(K=1)=0.296, TV(K=4)=0.256, bootstrap difference +0.040 with 95% CI [+0.029, +0.050] and one-sided p-value 0.0000. That established the architectural signal at the noiseless level. Hardware execution followed a passed pre-flight.
Run archived as QQ v3 evidence bundle. IBM Quantum Platform, SamplerV2 on IBM Heron r2 (ibm_fez). QPC noise reducer enabled: 3 runs per circuit, counts aggregated across runs. K-ablation numbers below are those aggregated counts.
| K | TVAB | TVBA | TVmean | KLmean |
|---|---|---|---|---|
| 1 (faithful non-polycontextural control) | 0.2815 | 0.3216 | 0.3015 | 0.4317 |
| 2 (intermediate transjunctional structure) | 0.2672 | 0.2491 | 0.2582 | 0.3296 |
| 4 (full polycontextural) | 0.2479 | 0.2541 | 0.2510 | 0.3115 |
Lower TV / KL means closer to the empirical Wang and Busemeyer joint distribution. Improvement is monotone across K on both metrics. The model fits empirical data more faithfully as polycontextural blocking is added, with parameters held fixed across K.
Each replicate resamples per-circuit counts from the multinomial defined by the observed shots, recomputes both TVmean values, and records the K=1 − K=4 difference.
| Metric | Mean K=1 | Mean K=4 | Mean diff | 95% CI of diff | One-sided p | Significant @95%? |
|---|---|---|---|---|---|---|
| Total-variation distance | 0.3015 | 0.2510 | +0.0505 | [+0.0359, +0.0655] | 0.0000 | YES |
| KL divergence | 0.4319 | 0.3119 | +0.1200 | [+0.0888, +0.1520] | 0.0000 | YES |
A one-sided p-value of zero from 2000 bootstrap replicates means every single resample showed K=4 fitting the empirical Clinton–Gore joints strictly better than K=1. The 95% confidence intervals on the difference sit entirely above zero on both metrics.
| Field | Value |
|---|---|
| Backend | ibm_fez (IBM Heron r2) |
| Mode | SamplerV2 on open-instance Runtime; readout proprietary readout |
| QPC noise reducer | Enabled; 3 runs per circuit aggregated; matrix readout mitigation applicable at 16 qubits |
| First / last job ID | d7v5p2jack5s73bf13jg (K=1 AB run 1) · d7v5q2nmrars73d7prsg (K=4 BA run 3) |
| Full job list | 18 IDs in ibm_job_ids_all |
On ibm_fez (IBM Heron r2), the QPC polycontextural architecture (K=4) reproduces the empirical Wang and Busemeyer Clinton–Gore joint-distribution shape strictly better than a faithful non-polycontextural control (K=1) of equal qubit count, depth, and shot budget. Bootstrap-significant at p<0.0005 on both total-variation and KL-divergence metrics. Parameters fit only from order-blind marginals. Joints held out for evaluation.
This is a quantitative architectural claim grounded in a controlled comparison against published empirical human data, with hardware-archived statistical significance, on a problem class whose classical theorem is already in the literature.
The K-ablation reports continuous statistical evidence: effect size, confidence interval, and p-value. TV(K=4) = 0.2510 is the closer fit to the published joints, with parameters held fixed across K (φAB and ψ taken from order-blind marginals). That is the architectural claim: polycontextural blocking does the work at matched resources. The 18 IBM job IDs are the hardware archive.
The published run is v3: TV / KL divergence against published Wang and Busemeyer empirical joints, simulator pre-flight (p=0.0000), then 18 SamplerV2 jobs on ibm_fez. Earlier metric work (QQ-residual) informed the choice of TV / KL as the within-quantum architectural discriminator. v1 hardware IDs remain in QQ v1 evidence bundle as an iteration record. v3 is the result this page reports.
Empirical target data: Wang and Busemeyer 2013, Topics in Cognitive Science 5(4):689–710, Table 1. PDF: https://jbusemey.pages.iu.edu/quantum/QuestOrdEff.pdf.
Numbers in this page come from:
QQ v3 evidence bundle: main hardware archive (this run)QQ v1 evidence bundle: earlier hardware archive (iteration record)