- Rating
- 1
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 1
- Soundness
- 1
Read review
Paper summary
The paper studies how symmetry constraints affect trainability and achievable accuracy in parameterized quantum models. It proposes a criterion linking gradient-variance scaling to the largest multiplicity-space dimension. Experiments examine gradient variance and final task errors under different degrees of alignment between architectural and task symmetry. The paper also investigates sensitivity to symmetry breaking and distinguishes approximation limits imposed by symmetry from errors due to incomplete optimization.
Critical strengths
S1. The motivation is clear. Understanding when symmetry facilitates optimization and when it excludes useful solutions is an important question for the design of equivariant models.
S2. The distinction between approximation and optimization provides a useful framework. The error decomposition in Section 3.2 separates the best performance achievable under a symmetry constraint from the remaining optimization error. This helps clarify why larger gradients alone do not guarantee better final performance.
Critical weaknesses
The concerns below are not exhaustive. I focus on the most significant weaknesses, which would need to be addressed before I could recommend acceptance.
W1. Theorem 1(i) requires nondegeneracy assumptions. For an -equivariant circuit with a permutation-invariant initial state, choose , the orthogonal projector onto the permutation-symmetric subspace. Although is polynomial in , the state remains in this subspace, giving and for every parameter. These equalities hold even under the blockwise 2-design assumption. The gradient variance is therefore zero, contradicting the claimed inverse-polynomial lower bound.
W2. Theorem 1(ii) is not valid under its stated assumptions. An exponentially large multiplicity block need not contribute to the loss. The gradient variance can remain constant even when grows exponentially.
Consider the following counterexample. Let , write , and represent the nonidentity element of by . The trivial and sign representations have multiplicities and , respectively, so . Choose , , and . Here, are the Pauli matrices, and the appended zeros denote a zero vector and a zero matrix of the indicated sizes. Initialize the remaining circuit parameters so that and are independent Haar-random unitaries. For every fixed , both blocks of remain Haar-distributed and therefore satisfy the exact unitary 2-design assumption. Moreover, commutes with , and the initial state is -invariant.
Since the initial state occupies only the first block, define . The cost is . Differentiation gives . Using , this becomes . The state is a Haar-random pure qubit. Each second moment is therefore , giving . This variance is independent of and contradicts the claimed upper bound.
W3. The orbit formula is inconsistent with the invariant initial-state assumption. Equations (6) and (14) include multiplicity blocks that an invariant pure state cannot occupy. For the usual permutation action of on qubits, the orbit has real dimension , contradicting the claimed scaling.
Let denote the unitary operator representing . Initial-state invariance means , while circuit equivariance means . Combining these identities gives , so every reachable state remains invariant. A nontrivial irreducible representation has no nonzero vector fixed by every group element. The reachable states therefore lie entirely in the trivial isotypic component, whose dimension is , the multiplicity of the trivial representation.
Under the full group of equivariant unitaries, the orbit consists of all normalized vectors in , with vectors differing only by a global phase identified. Thus , rather than the product stated in Equation (6). Its real dimension is : normalization and global phase each remove one degree of freedom from the real coefficients.
For qubits, permutation invariance requires equal amplitudes for all binary basis states with the same number of ones. An invariant state is therefore specified by one complex coefficient for each count , giving . Consequently, the full projective orbit is and has real dimension . A finite circuit can reach only a subset of this orbit, so its dimension is at most .
W4. The central 2-design assumption needs a clearer explanation. Section 3 introduces parameterized quantum states, expectation-value objectives, and equivariance. However, the unitary 2-design assumption underlying Theorem 1 is described only informally. The authors should explain what a 2-design is and how the circuit parameters are randomly initialized.
Questions for authors
See Weaknesses.