Question Bank

20 questions. Answers and explanations in solution-bank.md. No peeking.

Q1. What does the SamplerV2 primitive return for each PUB?

• A. Expectation values of observables supplied in the PUB
• B. Per-shot measurement samples / bitstring counts from the circuit's classical registers
• C. Quasi-probability distributions after readout mitigation
• D. The final statevector of the circuit

Q2. Given:

from qiskit import QuantumCircuit
from qiskit.primitives import StatevectorSampler
 
qc = QuantumCircuit(2)
qc.h(0); qc.cx(0, 1)
qc.measure_all()
result = StatevectorSampler().run([qc], shots=1024).result()
counts = result[0].data.c.get_counts()

What happens on the last line?

• A. counts is a dict like {'00': 5xx, '11': 5xx}
• B. AttributeError — the register is named meas, not c
• C. KeyError: 'c'
• D. An empty dict, because measure_all stores data under result.data.counts

Q3. Which is a valid SamplerV2 PUB that also fixes the shot count for that PUB?

• A. (circuit, shots)
• B. (circuit, observable, shots)
• C. (circuit, parameter_values, shots)
• D. (shots, circuit, parameter_values)

Q4. Given:

job = sampler.run([(qc, None, 333)], shots=1000)

How many shots are executed for this PUB?

• A. 1000 — the run() argument always wins
• B. 333 — per-PUB shots override the run()-level value
• C. 1333 — they are added
• D. It raises an error because shots are specified twice

Q5. Given:

result = sampler.run([qc_a, qc_b, qc_c]).result()

How do you get the counts of qc_b (all three use measure_all())?

• A. result.get_counts(qc_b)
• B. result[1].data.meas.get_counts()
• C. result.data[1].meas.get_counts()
• D. result[2].data.meas.get_counts()

Q6. A 3-qubit circuit built with measure_all() is run with shots=512. For ba = result[0].data.meas, what are ba.num_shots and ba.num_bits?

• A. num_shots=512, num_bits=3
• B. num_shots=3, num_bits=512
• C. num_shots=512, num_bits=8
• D. num_shots=1536, num_bits=1

Q7. What does result[0].data.meas.get_bitstrings() return?

• A. A dict mapping bitstrings to counts
• B. A list of unique bitstrings observed
• C. A list with one bitstring per shot (length = num_shots)
• D. A NumPy array of integers

Q8. Given:

q = QuantumRegister(2, 'q')
a = ClassicalRegister(1, 'a')     # added first
b = ClassicalRegister(1, 'b')     # added second
qc = QuantumCircuit(q, a, b)
qc.x(0)
qc.measure(0, a[0])               # always 1
qc.measure(1, b[0])               # always 0
res = StatevectorSampler().run([qc], shots=10).result()[0]
print(res.join_data().get_counts())

What is printed?

• A. {'10': 10}
• B. {'01': 10}
• C. {'1': 10, '0': 10}
• D. TypeError — join_data requires register names

Q9. Which line raises an error for a qiskit_ibm_runtime SamplerV2 instance?

• A. sampler.options.default_shots = 2048
• B. sampler.options.dynamical_decoupling.enable = True
• C. sampler.options.twirling.enable_measure = True
• D. sampler.options.resilience_level = 1

Q10. Which is a valid value for sampler.options.dynamical_decoupling.sequence_type?

• A. "XZXZ"
• B. "XpXm"
• C. "ZZ"
• D. "YY4"

Q11. Which set contains only real fields of sampler.options.twirling?

• A. enable_gates, enable_measure, num_randomizations, shots_per_randomization
• B. enable_gates, enable_measure, zne_factors, shots_per_randomization
• C. enable_twirling, randomization_count, measure_twirl
• D. enable_gates, resilience_level, num_randomizations

Q12. A colleague passes a circuit with no measurements to SamplerV2. Per the exam's framing, which statement is correct?

• A. Sampler measures all qubits automatically in the Z basis
• B. Sampler circuits must contain measurements — without them there is no data to sample
• C. Sampler infers the measurement from the observable in the PUB
• D. The job fails only on real hardware, but simulators add measurements

Q13. Given:

qc = QuantumCircuit(2, 2)
qc.h(0); qc.cx(0, 1)
qc.measure([0, 1], [0, 1])
result = sampler.run([qc]).result()

Which expression retrieves the counts?

• A. result[0].data.meas.get_counts()
• B. result[0].data.c.get_counts()
• C. result[0].get_counts()
• D. result[0].data.get_counts('c')

Q14. Your kernel restarted after submitting a job yesterday. You saved the job ID string. How do you reconnect to the job?

• A. job = QiskitRuntimeService().job("d1e2f3a4b5")
• B. job = SamplerV2.retrieve("d1e2f3a4b5")
• C. job = service.jobs("d1e2f3a4b5")
• D. You cannot — job handles die with the Python process

Q15. Which call correctly lists only your queued/running jobs on one backend?

• A. service.jobs(backend_name="ibm_torino", pending=True)
• B. service.jobs(backend="ibm_torino", status="active")
• C. service.list_jobs(filter="pending", device="ibm_torino")
• D. service.job(backend_name="ibm_torino", pending=True)

Q16. Which method returns the quantum processor execution time consumed by a job (in seconds), useful for tracking your account quota?

• A. job.metrics()
• B. job.usage()
• C. job.status()
• D. job.qpu_time

Q17. A 2-qubit Bell circuit returns counts = {'00': 412, '11': 588} over 1000 shots. What is the estimated probability of measuring '11'?

• A. 0.412
• B. 0.5 — Bell states are always 50/50 by definition
• C. 0.588
• D. 588

Q18. Single-qubit counts: {'0': 300, '1': 700} over 1000 shots. What is the manual estimate of ⟨Z⟩?

• A. +0.4
• B. −0.4
• C. 0.7
• D. −0.7

Q19. Your Sampler-based estimate of a probability has statistical error ε at 1000 shots. Approximately how many shots do you need to reduce the error to ε/2?

• A. 2000
• B. 4000
• C. 8000
• D. 1,000,000

Q20. Which statement about V2 result objects is TRUE?

• A. job.result() returns a PrimitiveResult, which is iterable and yields one PubResult per PUB
• B. job.result() returns a dict keyed by circuit name
• C. result[0].metadata holds the bitstrings; result[0].data holds shot counts only
• D. Job-level metadata is accessed as result[0].data.metadata