- Rating
- 2
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
The paper replaces the O(B²) pairwise fidelity objective in quantum embedding optimization with a prototype-based objective that needs O(B) fidelity evaluations. It derives a trace-distance lower certificate for class-ensemble distinguishability, an RKHS prototype classifier as a cheap witness for the soft-margin SVM, and a multi-prototype extension. Experiments on five small datasets show competitive QKSVM performance, with a clear gain on T-MNIST.
Critical strengths
-
The paper identifies a real bottleneck in data-adaptive quantum embedding optimization, namely the O(B²) pairwise fidelity evaluations repeated at every update, and addresses it with a change in objective structure rather than an approximation.
-
The paper reports its limitations and supports the central claims with diagnostic analysis rather than only headline numbers.
Critical weaknesses
-
Propositions 1–2 follow from standard trace norm inequalities, while Proposition 3 simply evaluates the SVM objective on a restricted classifier. The certificate is also vacuous on three of five datasets (Fig. 2), weakening its practical justification.
-
Pairwise-B samples pairs globally, whereas Full uses class-balanced averaging in Appendix D.1. Their weighting differs on imbalanced data. A stratified, class-balanced sampling baseline would make the comparison more convincing.
-
Linear scaling applies to fidelity counts per objective evaluation with fixed prototypes. Final QKSVM training still requires a quadratic kernel matrix. Total optimization costs, runtime, and finite-shot robustness are missing.
-
Table 1 reports means without uncertainty, and most improvements outside T-MNIST are small or inconsistent. Paired confidence intervals and additional data splits would help establish reliability.
-
The multi-prototype result is confounded. GLASS uses extra subclass labels and changes the scalarization strategy. Unsupervised prototypes and controlled objective ablations are needed to isolate the benefit of multiple prototypes.
Questions for authors
- Which configuration produced the GLASS and T-MNIST Prototype numbers in Table 1?
- Pairwise-B uses a refreshed random pair set per generation, while Prototype uses fixed anchors. Could authors report a fixed-pair budget-matched baseline, and ablate L_in, L_out, and L_sample separately? Without this, I cannot tell whether the T-MNIST gain comes from the certificate structure or from persistent anchors.?
- Are the test-split diagnostics in Section 3.2 and Appendices F.1 and F.2 used to make any design or reporting choice? The paper states that test data are not used for selection, but the margin and CKTA analyses are computed on test samples and support mechanistic conclusions.
6 comments
Ablation of and .
We first ablated the objective on T-MNIST, where the gain is clear. With one prototype per class, is identically zero, so the objective has two terms, and . We removed each term in turn, keeping the same fixed prototypes, GA budget and validation-based selection ( is Prototype variant, with paired 95% CI over 10 seeds):
| Objective | Test MCC | [95% CI] |
|---|---|---|
| (paper) | .761 | – |
| only | .006 | [, ] |
| only | .508 | [, ] |
Prototype objective without abilation outperforms both variants on every seed, and each term alone leads to a degenerate geometry. With only, all states collapse to one point and every fidelity is close to 1. With only, all states become mutually orthogonal, including each sample and its own prototype. As a result, the mean off-diagonal fidelity is and the effective rank of the kernel is 149 of 150, so the kernel approaches the identity. The two terms correspond to the two quantities in Proposition 1, prototype separation and sample-to-prototype distance, and neither suffices alone. Since both variants keep the same fixed prototypes, persistent anchors by themselves do not produce the T-MNIST gain.
Ablation of . is active only with several prototypes per class, so we ablated it in the multi-prototype GLASS setting of Section 3.3, removing each term in turn under the same GA budget:
| Objective | Test MCC | [95% CI] |
|---|---|---|
| (full) | .634 | – |
| without | .000 | [, ] |
| without | .620 | [, ] |
| without | .603 | [, ] |
| only | .552 | [, ] |
is also essential. Without it, all states collapse to one point, and the test MCC drops to zero. For the two attraction terms the picture is less sharp. Removing or alone changes the test MCC by and , and both intervals include zero. Removing both costs [, ]. We read this as partial redundancy between the two terms, which is consistent with how the prototypes are built. In proxy-based metric learning the class representatives are defined in the embedding space, either as free parameters trained by the loss (Proxy-NCA, Proxy-Anchor) or as cluster centers of the embedded points (magnet loss). Our prototypes are defined in the input space, as the circular mean of each sub-cluster, and are then embedded by the same circuit as the samples. A change of the circuit therefore moves a prototype together with the samples whose inputs lie close to that mean, and a term that acts only on prototypes still constrains these samples indirectly. The same reasoning explains why alone still reaches . This indirect control weakens as the inputs of a sub-cluster spread out around their mean, because samples far from the mean need not follow the prototype. Multi-center metric learning has the same structure, where guarantees on centers transfer to the examples only up to the sample-to-center distance (Qian et al., 2019). Controlling this distance directly is the role of . We keep both terms, since the full objective has the highest mean MCC, but we do not claim that its difference from the single-term variants is significant.
W3
Total optimization cost, runtime and finite-shot robustness; the final QKSVM still needs a quadratic kernel.
We agree. The linear scaling refers to the number of fidelities per objective evaluation, with the prototypes fixed before the search, and the final QKSVM still requires the training kernel and the Test Train kernel. These kernels are common to every objective and are computed once, after the optimization. The saving of the prototype objective lies inside the optimization loop, where the objective is evaluated for every candidate at every iteration. To measure both parts on a real QPU, we ran the full pipeline (GA and NQE) on IBM Heron processors (ibm_pittsburgh and ibm_boston) on T-MNIST (20 qubits, ) for ten seeds, with a training coreset of samples (20 per class). The same experiment is described in our answers to Reviewer LXe5 (W2, Q3) and Reviewer 9ftv (Q3), where we also give the per-step circuit counts for gradient-based training.
- Optimization. Each candidate is evaluated with inversion test at 1024 shots, and elite candidates are re-estimated with up to 4096 shots. The GA (population 50, 20 epochs per stage, 5 stages), together with 40 SPSA steps of NQE, took 10.6 QPU-hours per seed on average. An exact pairwise objective on the same samples needs circuits per candidate, 45 times as many.
- Final QKSVM kernel. After the optimization, the kernel of the selected embedding is measured once. The training kernel (upper triangle) and the Test Train kernel require 29,950 circuits per embedding at 1024 shots, about 2.4 QPU-hours per embedding (0.29 QPU-seconds per circuit, estimated from the submitted workload). This cost is quadratic in but is paid once and is the same for every objective. With the prototype objective it is smaller than the optimization cost above (10.6 QPU-hours). With an exact pairwise objective, the optimization would dominate the total cost by a wide margin.
- Performance of the hardware-optimized circuits. The circuits optimized entirely on hardware, with 1024-shot fidelity estimates and a coreset of 100 training subsamples, reach a Test MCC of with the hardware kernel. This is slightly below the noiseless Prototype result in Table 1 (), and remains above the noiseless Full pairs () and Pairwise-B ().
- Effect of shot noise on the kernel. Evaluating the same circuits with the noiseless kernel gives a Test MCC of , a difference of over ten seeds. The 1024-shot hardware kernels correlate with the noiseless kernels at –.
W4
Table 1 uncertainty. We now report the 10-seed mean standard deviation of Test MCC and ROC-AUC for every cell of Table 1. Table 6 also gives, for each baseline, a paired bootstrap 95% confidence interval of its difference from Prototype, with runs matched by seed.
On the four tabular datasets, most intervals include zero. After NQE, every tabular interval includes zero. At the GA stage, there are three exceptions:
- On IRIS, Prototype is above Pairwise-B ( [, ]) and marginally below Full pairs ( [, ]).
- On VERTEBRAL, Prototype is below Pairwise-B ( [, ]).
On T-MNIST, every interval excludes zero in favour of Prototype.
The fixed ZZ feature map scores below every other method on every dataset, and all of its intervals against Prototype exclude zero. To save space, we omit it from the table.
We agree with the reviewer that on the tabular datasets a random subsample of pairs performs as well as the prototype objective, and we no longer claim a gain there. The gain we now claim is restricted to T-MNIST, where Prototype outperforms every pairwise variant on all 10 seeds. We analyse where this gain comes from in our response to W2 and Q2.
W5
The multi-prototype result is confounded by the subclass labels and by the change of scalarization
We agree that the comparison in Section 3.3 changes several things at once, and we separated them with two controlled experiments on GLASS. Both keep the GA, its budget and the validation-based selection fixed.
Subclass labels. We replaced the sub-label prototypes by unsupervised ones. The prototypes are the -means centers of the inner training samples, and the cell weights are the cluster fractions. No subclass label is used. But the number of prototypes per class (), the objective and its scalarization are unchanged. The unsupervised prototypes reach a Test MCC of (10 seeds), compared with for the sub-label prototypes. Against the single-prototype objective of Table 1 () the difference is [, ], with of seeds higher, and against Full [, ]. Because both intervals reach zero, we describe the gain over a single prototype as consistent across seeds but not dramatic.
Objective terms. With the scalarization held fixed, we removed each term of the multi-prototype objective in turn (W2/Q2). The two cohesion terms, and , can replace each other, and alone does not account for the multi-prototype result. Its benefit comes from the additional prototypes.
Q1
Prototype configuration in Table 1
All Prototype rows of Table 1, including GLASS and T-MNIST, use a single prototype per class (, circular-mean prototypes). We set them to one because we want to check first our optimization works well. The multi-prototype GLASS results are reported separately in Section 3.3 and do not enter Table 1.
W1
Propositions 1–3 rely on standard inequalities, and the certificate is vacuous on three datasets.
We agree that Propositions 1–2 use standard trace-norm inequalities and that Proposition 3 evaluates the SVM objective at a feasible classifier. We do not claim a new norm inequality. Their contribution is to connect a small set of measurable prototype fidelities to the class-ensemble distinguishability ΔΦ and to the downstream SVM objective. ΔΦ is a trace norm between two mixed states and cannot be measured directly, so existing embedding objectives replace it with a Hilbert–Schmidt relaxation over all B(B−1)/2 pairwise fidelities (Lloyd et al., 2020; Hur et al., 2024). Because the prototype states are pure, Propositions 1–2 bound ΔΦ itself from B+K(K−1)/2 fidelities, and Proposition 3 bounds the QKSVM objective from the same fidelities, without forming the B×B kernel.
The certificate is a measurable lower bound on ΔΦ, obtained from B+K(K−1)/2 fidelities. On IRIS and T-CELL the low-loss circuits reach a positive value (Figure 2(a)), so there it certifies a non-trivial class-ensemble distinguishability at linear cost. On the other three datasets it is negative, and hence vacuous as a lower bound. Even there it keeps its second role, which is to determine which quantities the objective should contain and how they should be combined. Two observations support this role.
Both terms of the bound are needed. Removing either term of the objective on T-MNIST leads to a degenerate embedding (Q2). With the sample-to-prototype term alone, all states collapse to one point (Test MCC 0.006). With the prototype-separation term alone, all states become mutually orthogonal and the kernel approaches the identity (Test MCC 0.508, against 0.761 for the full objective, all ten seeds).
The certificate tracks training performance on the tabular datasets, even where its value is vacuous. Figure 2(c) reports, over the circuits of Figure 2, the correlation of the Train-CV MCC of the QKSVM with the certificate (column "Prototype") and with the negative full pairwise loss (column "Full"). On the four tabular datasets the correlation with the certificate is positive (r=0.40–0.66), including VERTEBRAL and GLASS, where the certificate is negative. On noiseless T-MNIST it is negative (r=−0.39), but the full pairwise loss on the same circuits gives the same value (r=−0.39), so this behaviour is not specific to the prototype certificate. On IBM Heron, where the certificate of the selected circuit of each seed is estimated from measured counts, the correlation is positive (r=0.77, ten seeds).
W2 and Q2
A class-balanced, budget-matched fixed-pair baseline, and a separate ablation
Fixed-pair baseline. Following the reviewers' suggestion (and Reviewer 9ftv), we added a pairwise baseline with fixed, class-balanced pairs on every dataset. It uses the same number of fidelity evaluations per step as Prototype (e.g., 160 pairs on T-MNIST). Half of the pairs are same-class pairs, the classes and class pairs are sampled uniformly, and the pairs are drawn once and kept fixed for the whole run. Results are in Table 6.
On T-MNIST, Prototype outperforms all baselines on all 10 seeds:
Class balancing does not close the gap. The fixed class-balanced baseline scores even below the original Pairwise-B.
The difference in weighting between Pairwise-B and Full therefore does not change any conclusion. On T-MNIST, persistence alone does not explain the gain. The fixed pairs are persistent, yet they are below Prototype on every seed and also below the refreshed Pairwise-B. Appendix G.1 shows that the fixed-pair objective already separates the training samples weakly, and that its fit is tied to its own pairs (loss 0.175 on the 160 optimized pairs, 0.277 on all pairs of the same samples).