One polycontextural layer on IBM Heron r2, QuEra Aquila, and the vendor fleet.
Co-resident contextures, one job, public IDs, architecture metrics on raw counts.
CO₂ multi-context · Torino → · Mask synthesis · Kingston → · ADMET · Marrakesh → · Real-World Desks → · Pharma Atlas → · Preprint: Zenodo
Architecture desks on named hardware. Raw counts. Public job IDs.
What: Eight climate logics on the top 20 emitters. OWID CO2 and World Bank GDP in the encoding.
Result: Eight public jobs on IBM Eagle · ibm_torino. Logical width 520. Entropy 13.0. First job d65hqspv6o8c73d30o0g.
What: ICCAD-2013 photomask ladder. Six masks, 24 measured contextures, one job on ibm_kingston.
Result: Raw coupled ranking: 0.622 (no-OPC) ahead of ILT 0.460. Job daas1burrl7c73860bd0.
What: Seventeen ADMET logics, locked junction RYY, 156 qubits on ibm_marrakesh.
Result: Raw junction gap 0.185 · retention 0.749 · separable 0.184. Same-day full pass.
ADMET desk →What: Frustrated magnet as three co-resident contextures in one circuit.
Result: 128Q campaign-mean ICC 0.187. ctx2 |ZZ| 0.101 against the single-context control at 0.285.
PFQM →What: Coupled job versus separable and intracontext controls on ibm_fez.
Result: L1–L5 pass on raw ICC. Gaps 0.289 · 0.101 · 0.130. Intracontext 3/3.
Open challenge →What: One cyber desk matrix on ibm_fez (156Q) and QuEra Aquila (153 atoms), plus a 256-atom full-register run.
Result: Architecture PASS on Aquila. IBM job d9eephkjeosc73fimti0. 256-atom |Δ ZZ_bridge| 0.283.
What: Gauge-sector priors on the JSC ICC stack. Even / odd / bridge contextures.
Result: July 2026 · 3/3 raw pass · gaps 0.349 · 0.243 · 0.131.
MCGS report →What: Irreducible 156-asset portfolio as one polycontextural IBM job.
Result: QPC 7.587 · Classiq-cut 9.072 · job d9b7ki7u62qs738ot6ig.
What: One co-resident desk on NVIDIA CUDA-Q 0.15.1, IBM Heron as the QPU arm.
Result: 3→1 jobs · Kingston d9tp5ic98n5s7391u4m0.
What: Cerrado ESG, Beijing traffic, insurance portfolio. One polycontextural submission each on ibm_fez.
Result: 156Q layout, auditable IDs, plain-language scores on each pilot.
Trilogy →What: Grid, polypharmacy, organizing, fair lending, EO tasking, plus mask synthesis.
Result: Multi-logic decisions on IBM Heron with public jobs.
Desks →What: Wang–Busemeyer question-order data. Four contextures versus one, same 16 qubits and shots.
Result: K=4 beats K=1 on held-out joint prediction. 18 IBM IDs. Bootstrap p < 0.0005.
QQ report →What: Orchestrator on IBM Open / Heron 156Q. Four registers to 24 contexts.
Result: 13 auditable IBM jobs. JSON schema qpc_benchmark_v1.
Scale table →What: Range, safety, and payload as co-resident contextures on ibm_fez. Full 156-qubit Heron r2 register, one job.
Result: Coupled ranking from that sitting. Public final report with Fez layout figures.
Aerospace report →Vendor ports, sampling tests, and domain pilots with published IDs. Each link is a full report.