Peaked Quantum Circuits (HQAP) is a Parquet collection of OpenQASM sampling benchmarks engineered to exhibit heuristic quantum advantage while remaining classically verifiable. Derived from the Quantum Advantage Tracker’s classically-verifiable-problems repository and the study “Heuristic Quantum Advantage with Peaked Circuits” (Gharibyan et al., arXiv:2510.25838), the dataset packages raw QASM together with structural metadata for scalable analytics, benchmarking, and ML workflows.

Peaked circuits are designed so that a single target bitstring dominates the output distribution, creating strong statistical signatures of quantum interference with relatively shallow sampling. Heuristic quantum advantage refers to regimes where quantum processors outperform classical heuristics on structured—but not worst-case—problems. Peaked circuits sit squarely in this regime by amplifying interference patterns that are difficult for classical simulators to reproduce efficiently.

Crucially, the target bitstring and its probability can be computed and verified efficiently on classical hardware. This makes each circuit suitable for classically verifiable quantum advantage experiments, while still stressing modern simulators, tensor-network methods, and hardware calibration pipelines.

Each row in the Parquet file corresponds to one HQAP circuit instance from the Quantum Advantage Tracker. The dataset includes circuit families such as peaked_circuit_heavy_hex and peaked_circuit_P9, typically spanning 49–56 qubits with hundreds to a few thousand entangling gates.

The dataset stores the full OpenQASM source, declared qubit and gate counts, and optional peak metadata so users can filter by circuit size, connectivity, or verification parameters without repeatedly parsing source files.

Columns capture identifiers, hardware-agnostic circuit metrics, optional verification metadata, full QASM text, and provenance back to the original repository. Nullable fields allow for future HQAP variants without breaking schema compatibility.

Polars

import polars as pl

from aqora_cli.pyarrow import dataset

df = pl.scan_pyarrow_dataset(

    dataset("aqora/quantum-advantage-tracker-peaked-circuits", "v0.0.0")

).collect()

print(df.head())

Pandas

import pandas as pd

df = pd.read_parquet(

    "aqora://aqora/quantum-advantage-tracker-peaked-circuits/v0.0.0"

)

print(df.head())