P12 peak recovery on Quantinuum Helios-1 #247

Name

Quantinuum Helios-1 recovery for peaked_circuit_P12_Hqap_98x2457

Circuit

peaked_circuit_P12_Hqap_98x2457

Value

100

Method

Quantinuum Helios-1 hardware with frequency, Hamming-space medoid, and cluster-consensus reconstruction

Method proof

Summary

We executed the existing tracker instance

peaked_circuit_P12_Hqap_98x2457

on Quantinuum Helios-1 hardware.

The run requested 200 shots and reported a cost of 1373.4 HQC. The reconstructed sample contained 198 unique bitstrings. The most frequent string, weighted observed medoid, and cluster consensus all recovered the same 98-bit candidate.

The most frequent string is the observed string with the highest reconstructed-shot count. The weighted observed medoid is the observed string minimizing the total Hamming distance to the reconstructed sample, with repeated strings weighted by their observed frequencies. Cluster consensus identifies the dominant group of nearby strings and applies a bitwise majority vote within that group.

All three methods recovered the following 98-bit string:

10100011110010100111000100011100110001011111011100111001010110101011001001000000101000100000111100

The result was externally accepted by the BlueQubit Peaked Portal. The hidden target was not used during the reconstruction procedure; it was used only for subsequent external verification.

Reproducibility package

The complete source, raw provider result, repaired framing record, reconstructed shots, metadata, timing records, analysis code, figures, and SHA-256 manifests are available here:

https://github.com/MonitSharma/peaked_circuits/tree/main/results/tracker_submissions/p12

The main repository is available here:

https://github.com/MonitSharma/peaked_circuits

Provider job ID:

d5cba0df-a7aa-459e-ac51-8092645c059f

Canonical P12 QASM SHA-256:

868ff86a396f86a8cbca48f7127e49c4c4f951a5a955295e5010a92b76be961d

Resource details

• Provider: Quantinuum
• Hardware: Helios-1
• Requested shots: 200
• Reconstructed shots analyzed: 200
• Reported HQC cost: 1373.4
• Nested provider execution time: 237.588 seconds
• Submitted-to-completed provider time: 4580.972377 seconds
• Outer running-to-completed provider time: 4160.855590 seconds
• Classical analysis time: 0.010313208 seconds
• Classical resource: Apple Silicon Mac

The provider timing fields are reported separately because they represent different stages of the job lifecycle. The classical runtime covers input loading, validation, count construction, most-frequent-string recovery, weighted-observed-medoid recovery, and cluster-consensus recovery. It excludes Python interpreter startup, plotting, provider queue time, and provider execution.

External verification

BlueQubit Peaked Portal:

https://app.bluequbit.io/hackathons/oEOtLSSrPSVH60Ah?tab=leaderboard

Leaderboard evidence:

https://github.com/MonitSharma/peaked_circuits/blob/main/docs/assets/bluequbit_leaderboard_20260918.png

Quantum runtime (seconds)

237.588

Classical runtime (seconds)

0.010313208

Compute resources (quantum)

Quantinuum Helios-1

Compute resources (classical)

Apple Silicon Mac; local deterministic Python post-processing

Notes

200 shots; framing repaired; no advantage claim

Authors

Monit Sharma, Hoong Chuin Lau

Institutions

Singapore Management University