- Rating
- 3(+1)
- Confidence
- 1(−2)
- Informativeness
- 1
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This is a timely and technically interesting paper. The central idea is to learning a surrogate that ranks SQD sampling circuits from circuit structure, Hamiltonian descriptors, and low-shot outcomes. The key step is a successive-halving/evolutionary search, which is clearly described. The experiments show consistent improvements over vanilla SQD and low-shot-guided QAS. On TFIM and XXZ benchmarks, it reports strong ranking correlations, generalization to nine unseen 20-qubit TFIM Hamiltonians, and lower SQD errors than several baselines. However, the strongest scalability claims rely on per-target retraining with very large offline high-shot datasets, so the practical amortized advantage is not yet established.
Critical strengths
The family-level formulation targets an important SQD bottleneck and is conceptually novel (but there are multiple surrogate based quantumn circuit design works as of 2026). The multimodal design is sensible; ablations support the contributions of measurements, circuit information, and Hamiltonian context.
Critical weaknesses
Offline cost is substantial and excluded from headline budget comparisons. Training labels use roughly one million shots per circuit, reaching billions of simulated shots; the paper does not quantify how many target problems are needed to amortize this cost. The 10 to 40-qubit scaling, VQE, noise, and molecular studies train separate predictors for each target Hamiltonian. Thus, the strongest scalability results demonstrate transfer across circuits, not the central claim of transfer across Hamiltonian. All evidence is simulation-based. The noise study retrains on the tested noise levels, leaving robustness to real quantunm device drift or unseen noise unclear.
Several comparisons use only five seeds and then three for ablations, all comes without statistical tests. Some improvements over low-shot QAS are modest.
Questions for authors
What is the full offline-plus-online shot and classical-compute cost, and when does QAS2QD become cheaper than instance-wise search?
Can one predictor trained across Hamiltonians reproduce the 30/40-qubit, molecular, and noisy results?
How does performance change under unseen noise or real-hardware drift?
Are the conclusions robust across more seeds and paired significance tests?
8 comments
We thank the reviewer for the constructive feedback on resource costs, Hamiltonian transfer, noise robustness, and statistical reliability. In response, we clarify offline-cost amortization, evaluate shared predictors on held-out Hamiltonians, and provide additional evidence from unseen-noise tests and experiments on a real quantum processor. We also expand the multi-seed evaluation and add paired significance tests to assess statistical robustness. We will incorporate these clarifications and additional results into the revised manuscript.
W1-1: Offline cost is substantial and excluded from headline budget comparisons. Training labels use roughly one million shots per circuit, reaching billions of simulated shots; the paper does not quantify how many target problems are needed to amortize this cost. Q1: What is the full offline-plus-online shot and classical-compute cost, and when does QAS2QD become cheaper than instance-wise search?
A: We thank the reviewer for highlighting the need to account for the offline investment. Below, we quantify this cost and clarify the conditions under which it can be amortized by reusing the same trained predictor across target problems.
(1) Measurement cost and amortization. Our family-level formulation separates a one-time offline investment from the measurement costs incurred for each target Hamiltonian. We compare against instance-wise high-shot SQD-guided search, which evaluates candidate circuits with high-shot measurements separately for each target. The offline dataset contains 2,000 source circuits measured with shots each. The same records can support multiple Hamiltonians and low-shot training inputs without additional measurements. During online search, predictor guidance scores active candidates using substantially fewer shots than the high-shot reference, which uses the same per-circuit shot budget as training-label estimation ( shots). The table below summarizes the measurement costs for the 20-qubit TFIM example.
| Component | Predictor-guided method | Instance-wise high-shot search |
|---|---|---|
| Offline acquisition shots, once per predictor | ||
| Search-guidance shots per target | ||
| Final-evaluation shots per target | ||
| Total measurements per target (search + final evaluation) |
For target Hamiltonians, the total costs are and . Here, denotes the per-target measurement cost, including circuit search and final SQD evaluation, while denotes the one-time offline acquisition cost. Both methods independently evaluate five final circuits with shots each; their final errors are 0.198 and 0.486, respectively. Using the selected-dataset acquisition cost lower bound gives $ M>\frac{C_{\mathrm{offline}}}{C_{\mathrm{online}}^{\mathrm{high}}-C_{\mathrm{online}}^{\mathrm{pred}}}=\frac{2\times10^9}{1.349\times10^9-9.242\times10^6}\approx1.49. $ This yields an optimistic conditional projection of two target Hamiltonians, assuming the same frozen predictor continues to attain the required accuracy under the same search budget. Additional offline acquisition costs can increase the required number of targets.
(2) Classical computation. The classical costs and their frequency are summarized below.
| Component | Classical computation | Frequency |
|---|---|---|
| Offline labeling | Subspace construction and SQD diagonalization for circuit–Hamiltonian pairs | Once per offline dataset |
| Predictor training | Neural-network optimization | Once per predictor |
| Circuit search | Evolutionary operations and candidate selection | Per target search |
| Candidate evaluation | Low-shot SQD diagonalization and predictor inference | Per active candidate at each search stage |
| Final SQD calculation | Subspace construction and diagonalization | Per final evaluation |
Relative to direct SQD-guided search, predictor guidance adds offline label construction, neural-network training, and recurring inference costs. Offline labeling and training are shared across target Hamiltonians, while search and final evaluation recur for each target.
W1-2: The 10 to 40-qubit scaling, VQE, noise, and molecular studies train separate predictors for each target Hamiltonian. Thus, the strongest scalability results demonstrate transfer across circuits, not the central claim of transfer across Hamiltonian. Q2: Can one predictor trained across Hamiltonians reproduce the 30/40-qubit, molecular, and noisy results?
A: We thank the reviewer for raising this distinction. We revisited the manuscript’s scaling, molecular, and noisy experiments by changing the predictor-training regime from target-specific training to joint training across Hamiltonians. The original model architecture, circuit space, target test datasets, initial populations, , search and evaluation procedures, measurement budgets, and paired search seeds remain unchanged. Shared predictors are frozen before target evaluation. Target-Hamiltonian labels are excluded from fitting, normalization, and checkpoint selection.
The three studies are reported separately below. Entries give mean sample standard deviation of final SQD energy errors with five paired search seeds in units of .
1. Scaling from 10 to 40 qubits. At each system size, one predictor is trained across , validated at , and evaluated at the held-out , with .
| Qubits | Target-specific predictor | Shared predictor across Hamiltonians |
|---|---|---|
| 10 | ||
| 20 | ||
| 30 | ||
| 40 |
2. Molecular benchmark. For 16-qubit N$_2$ with , one predictor is trained across six source geometries, Å, validated at Å, and reused at the three held-out bond lengths below.
| Held-out bond length | Target-specific predictor | Shared predictor across Hamiltonians |
|---|---|---|
| Å | ||
| Å | ||
| Å |
3. Noisy benchmark. For the noisy 20-qubit XXZ experiment, we use the same Hamiltonian split as in the scaling study and retain the original noise levels and . One predictor is trained jointly across the source Hamiltonians and noise levels, then frozen for evaluation at the held-out .
| Noise strength | Target-specific predictor | Shared predictor across Hamiltonians |
|---|---|---|
Yes: within each setting, a predictor trained across source Hamiltonians can be reused for the held-out targets without target-specific retraining, with one predictor per system size in the scaling study. These results extend the evidence for Hamiltonian transfer to the full 10–40-qubit scaling study, molecular geometries, and noisy circuits. We will replace the manuscript’s target-specific results in these three studies with the shared-predictor results, update the corresponding VQE comparison, and document the joint-training protocol and Hamiltonian splits.
I have done a more careful reading through this article given the conflicting review scores; I have to say that I tend to agree that point 3 of B3vd is a critical flaw, and I cannot see how reviewer B3vd's concern can be addressed convincingly without more extensive experiments.
In addition, I did a quick search in surrogate-based circuit design papers recently posted (e.g., https://arxiv.org/abs/2510.00145 and https://arxiv.org/abs/2507.12256). I would tend to suggest that authors can expand these literature, make a more comprehensive comparison to make the contribution more solid. I do hope my comment could encourage authors to make this interesting work more evidenced.