IBM C1000-179 — Fundamentals of Quantum Computing Using Qiskit v2.X Developer Passing target: 24/34. All code outputs were verified by execution with qiskit 2.5.1 / qiskit-ibm-runtime 0.48.0 / qiskit-aer.

Answer Key

Q	Ans	Q	Ans	Q	Ans	Q	Ans
1	A	10	B	19	A	28	A
2	A	11	B	20	B	29	B
3	D	12	A, B	21	A	30	B
4	A	13	A	22	A	31	A
5	A	14	B	23	A	32	A
6	B	15	A	24	A	33	A
7	A	16	B	25	B	34	A
8	B	17	B	26	A
9	A	18	A	27	B

Explanations

Q1 — A ({'11': 100}). x(0) guarantees the mid-circuit measurement stores 1, so the if_test branch always fires and flips qubit 1. Both clbits end up 1: little-endian bitstring '11'. Verified by execution. Distractor analysis (tricky #1): B is for readers who think the if_test never triggers (forgetting the X before the measurement); C/D confuse bit ordering or imagine randomness where the circuit is fully deterministic.

Q2 — A ({'1': 100}). If the mid-circuit measurement gives 0, the if branch applies X → final measurement 1. If it gives 1, the else branch applies Z, which does not change |1⟩ → final measurement 1. Every shot ends in '1' — this is the active-reset-style pattern inverted. Re-measuring the same clbit simply overwrites it (D false). Verified by execution. Distractor analysis (tricky #2): C is the trap for stopping at "H gives 50/50" without following both branches to the same endpoint.

Q3 — D. if_test, while_loop, for_loop, and switch are the QuantumCircuit control-flow builders. There is no goto in Qiskit or OpenQASM 3.

Q4 — A (['a', 'b']). qc.parameters is sorted alphabetically by name, not by insertion order — even though b was used first. This matters when binding with a plain list in assign_parameters or in a PUB. Verified by execution. Distractor analysis (tricky #3): B (insertion order) is the classic list-binding bug that silently swaps your angles.

Q5 — A. Convert the sub-circuit with to_gate(), then call .control(1) to add one control qubit, and append it to [control, target]. controlled(), qc.control(...), and to_gate(control=...) do not exist.

Q6 — B ({'10': 1.0}). a.tensor(b) builds a ⊗ b with b occupying the lower qubit indices: qubit 0 stays |0⟩ (from b) and qubit 1 gets the X (from a), giving '10'. Verified by execution. Distractor analysis (tricky #4): A assumes a lands on qubit 0 because it is written first — the tensor-order/index-order mismatch is a favorite exam trap.

Q7 — A (1.0). Two little-endian rules stack: from_label('10') puts q1 = |1⟩, q0 = |0⟩, and Pauli('IZ') puts Z on qubit 0 (rightmost character). Z acting on |0⟩ gives +1. Verified by execution. Distractor analysis (tricky #5): B (−1.0) is the double-trap answer — reading either the label or the Pauli big-endian pairs Z with the qubit in |1⟩.

Q8 — B (ZX). A.tensor(B) is A ⊗ B with B on the lower qubit indices, so X sits on qubit 0 — the rightmost label character — giving 'ZX'. C confuses tensor with the matrix product (ZX ∝ iY); no phase appears in a tensor product. Verified by execution.

Q9 — A (ZIIZ). from_sparse_list pairs the k-th character with the k-th index: Z→qubit 0, Z→qubit 3. In the label, qubit 3 is leftmost and qubit 0 rightmost: ZIIZ. The whole point of the sparse form is that the string only covers listed qubits (D false). Verified by execution. Distractor analysis (tricky #6): B (ZZII) reads the indices as string positions from the left.

Q10 — B (SparsePauliOp('ZII')). Little-endian labels list qubit n−1 first: for qubit 2 of three, Z is the leftmost character. A ('IIZ') targets qubit 0 — the mirror-image trap. Observables carry no runtime qubit index (D).

Q11 — B ({'0': 0.5, '1': 0.5}). Tracing out one half of a Bell pair leaves the maximally mixed single-qubit state I/2 — measurement outcomes 0 and 1 are equally likely. This is the algebraic counterpart of the "zero-length Bloch vector" picture. Verified by execution.

Q12 — A, B. X and Z anticommute (A true, C false), and rz(π) = diag(e^{−iπ/2}, e^{iπ/2}) = −i·Z, i.e., Z up to global phase (B true). Y = iXZ, so D is off by the crucial i. Distractor analysis (tricky #7): D is tempting because "Y is made of X and Z" is half-remembered; forgetting the phase factor i breaks Hermiticity of the product.

Q13 — A. service.least_busy(operational=True, simulator=False) is the qiskit-ibm-runtime API. C is the retired provider-era helper function; B and D are invented.

Q14 — B (5). FakeManilaV2 mimics the 5-qubit Manila device; fake backends carry real device snapshots for local, noise-model testing. Verified by execution.

Q15 — A. Passing a backend object as mode= puts the primitive in job mode. There is no "job" string mode (B), no session= kwarg (C), and backend.run is the removed V1 pathway (D).

Q16 — B. Dynamic circuits are still circuits: they must be transpiled to the target ISA like everything else, and V2 primitives on supporting backends execute the classical feedforward. A, C, and D are all false.

Q17 — B (128). The third PUB element is a per-PUB shot count that overrides the job-level shots=512 for that PUB. Verified by execution. Distractor analysis (tricky #8): A is for assuming the run-level argument always wins; C adds them.

Q18 — A ({'1': 50}). Each classical register appears as its own attribute on the DataBin — data.alpha holds only the 1-bit register measuring qubit 0, which X set to 1. join_data() exists for combining registers but is not required to read one (D false). Verified by execution.

Q19 — A. join_data() concatenates the BitArrays (or arrays) of all classical registers of that PUB result into one object, useful when a circuit measured into several registers but you want unified bitstrings.

Q20 — B (3 100). The BitArray has num_bits = 3, and every shot reads '100': qubit 2 was flipped and occupies the leftmost bit position. A is the endianness flip; C confuses num_bits with the number of possible outcomes (2³ = 8). Verified by execution.

Q21 — A. Twirling options live under options.twirling with enable_gates and enable_measure booleans. B is Estimator resilience (not a Sampler option), C assigns a string to an options group, and D is dynamical decoupling.

Q22 — A ([ 1. -1.]). ry(0) leaves |0⟩ (⟨Z⟩ = +1); ry(π) rotates |0⟩ to |1⟩ (⟨Z⟩ = −1). One PUB with two parameter sets broadcasts to two expectation values in that order. Verified by execution.

Q23 — A ([-1. 1.]). After x(0), qubit 0 is |1⟩ and qubit 1 is |0⟩. 'IZ' measures Z on qubit 0 (rightmost char) → −1; 'ZI' measures Z on qubit 1 → +1. Verified by execution. Distractor analysis (tricky #9): B is the pure big-endian reading — the single most common mistake on the real exam per IBM's own trap list.

Q24 — A. The ISA circuit and a layout-mapped observable travel together in the PUB: (isa, obs.apply_layout(isa.layout)). The service does not remap observables (C), and the Estimator does not transpile for you (B).

Q25 — B. Estimator circuits must be measurement-free: the observable determines the measurement bases, and the primitive returns ⟨ψ|O|ψ⟩ estimates, not bitstrings. Passing a measured circuit is an error, not a no-op (C).

Q26 — A. draw() always puts q_0 on the top wire, so x(0) is the top box. B is the "physical intuition" flip; there is no version-dependent default for reverse_bits (D) — it is False unless you ask. Diagram generated with qiskit 2.5.1.

Q27 — B (−Y). The Y-basis states are |±i⟩ = (|0⟩ ± i|1⟩)/√2: plus at +Y, minus at −Y. The X axis belongs to |±⟩ and the Z axis to |0⟩/|1⟩.

Q28 — A. plot_bloch_multivector takes a state (Statevector or DensityMatrix) and renders one Bloch sphere per qubit. plot_bloch_vector takes a single [x, y, z] vector, not a circuit; the other two calls are invented.

Q29 — B. GHZ = (|000⟩ + |111⟩)/√2, so ideal sampling gives only 000 and 111, each near 50%. Equal height across all eight outcomes (A) is the signature of H on every qubit, not GHZ.

Q30 — B. job.done() returns a boolean immediately. job.result() blocks until the job completes; job.metrics() reports timing/usage metadata, not completion state; there is no job.wait().

Q31 — A (−0.2). Qubit 2 is the leftmost bit of each key. '000' → +1 (400 shots), '111' → −1 (600 shots): ⟨Z⟩ = (400 − 600)/1000 = −0.2. Distractor analysis (tricky #10): B flips the sign by mis-assigning which outcomes count as +1; C forgets the weighting; D is just p(111).

Q32 — A. job.usage() reports the quantum time (QPU seconds) the job consumed — the number that counts against your account's quota. Shots and PUB counts live elsewhere in metadata.

Q33 — A. The exporter emits c[0] = measure q[0]; followed by if (c[0]) { x q[0]; } — OpenQASM 3's C-style conditional block. Classical feedforward is precisely what OpenQASM 3 added over version 2 (D backwards). Verified by execution.

Q34 — A (2 2). qiskit.qasm2.loads still ships in Qiskit 2.x for interoperability (C false). measure q -> c; on a 2-bit register expands to two measure instructions, and qreg q[2] gives 2 qubits. Verified by execution.

Ten Trickiest — Quick Index

#	Q	Trap theme
1	Q1	Assuming the `if_test` branch may not fire
2	Q2	Not tracing both `if`/`else` branches to the same outcome
3	Q4	`qc.parameters` alphabetical vs insertion order
4	Q6	`tensor()` operand-vs-qubit-index ordering
5	Q7	Double little-endian: state label AND Pauli string
6	Q9	`from_sparse_list` index-to-character mapping
7	Q12	Y = XZ "identity" missing the factor i
8	Q17	Per-PUB shots overriding job-level shots
9	Q23	`'IZ'` vs `'ZI'` big-endian misreading
10	Q31	Sign of ⟨Z⟩ computed from counts